cs.RO / 1 / 2607.15330
Xiaomi-Robotics-1: Scaling Vision-Language-Action Models with over 100K Hours of Real-World Trajectories
小米机器人1：利用超过10万小时的真实世界轨迹扩展视觉-语言-动作模型
Xiaomi Robotics Team, Guo, Jun, Jin, Piaopiao, Li, Jason, Li, Peiyan, Li, Yingyan, Liu, Futeng, Peng, Wanli, Qin, Optimus, Su, Yifei, Sun, Nan, Sun, Qiao, Suo, Runze, Wang, Heyun, Wang, Yunhong, Wu, Rujie, Xia, Caoyu, Zhang, Lina, Zhao, Jack, Chen, Guoliang, Chen, Wenlong, He, Xinze, Li, Bin, Li, Qing, Li, Zhuorong, Qu, Heng, Song, Wenxuan, Xiang, Diyun, Xie, Yifan, Xu, Peiran, Ye, Hangjun, Ye, Wen, Zhao, Han, Zhou, Quanyun
Abstract
We present Xiaomi-Robotics-1, a foundational vision-language-action (VLA) model capable of (1) following diverse language instructions to perform a wide range of mobile manipulation tasks in unseen environments out-of-the-box, and (2) efficiently adapting to novel downstream tasks with minimal fine-tuning data. We propose a two-stage training recipe consisting of pre-training and post-training. During pre-training, we imbue the model with broad and generalizable action-generation capabilities by training on over 100k hours of real-world manipulation trajectories collected via UMI devices. Crucially, we develop a scalable auto-labeling pipeline that annotates trajectory clips with natural languages describing scene state transitions, providing rich and precise conditioning for action learning. During post-training, we aim to align these capabilities with robot embodiments and imperative instructions that humans naturally use to prompt robots. Extensive experiments demonstrate strong scaling behavior. Xiaomi-Robotics-1 consistently improves with increased data scales and model sizes during pre-training. This scaling behavior directly transfers to post-training, where a stronger pre-training model yields better out-of-the-box real-robot performance in unseen environments. Furthermore, Xiaomi-Robotics-1 serves as a strong robot foundation policy that can be efficiently fine-tuned on complex, dexterous tasks with high data efficiency. Across multiple simulation benchmarks, Xiaomi-Robotics-1 outperforms state-of-the-art methods. Notably, it establishes a new state-of-the-art with a 57.6% success rate on RoboCasa365, surpassing the previous best of 46.6%. Furthermore, it achieves an average score of 20.07 on RoboDojo, significantly outperforming the prior state-of-the-art (13.07). Code and model checkpoints will be released. Project page: https://robotics.xiaomi.com/xiaomi-robotics-1.html
Chinese Translation
我们提出了小米机器人1，这是一个基础的视觉-语言-动作（VLA）模型，能够（1）根据多样的语言指令在未见环境中执行广泛的移动操作任务，并且无需额外调整，和（2）以最小的微调数据高效适应新的下游任务。我们提出了一种包含预训练和后训练的两阶段训练方案。在预训练阶段，我们通过在超过10万小时的通过UMI设备收集的真实世界操作轨迹上进行训练，使模型具备广泛且可泛化的动作生成能力。关键是，我们开发了一个可扩展的自动标注管道，用自然语言对轨迹片段进行注释，描述场景状态的转变，为动作学习提供丰富而精确的条件。在后训练阶段，我们旨在将这些能力与机器人形态和人类自然使用的指令进行对齐，以提示机器人。大量实验表明了强大的扩展性。小米机器人1在预训练过程中，随着数据规模和模型规模的增加，性能持续提升。这种扩展性直接转移到后训练阶段，其中更强的预训练模型在未见环境中提供更好的真实机器人性能。此外，小米机器人1作为一个强大的机器人基础策略，可以在复杂的灵巧任务上高效微调，具有较高的数据效率。在多个仿真基准测试中，小米机器人1的表现超越了最先进的方法。值得注意的是，它在RoboCasa365上以57.6%的成功率创下了新的最先进记录，超过了之前的最佳记录46.6%。此外，它在RoboDojo上的平均得分为20.07，显著超越了之前的最先进记录（13.07）。代码和模型检查点将会发布。项目页面：https://robotics.xiaomi.com/xiaomi-robotics-1.html
cs.RO / 2 / 2607.15395
NeuroCommitSSM: Decision-Centric Shared Autonomy for Safe Assistive Manipulation via EEG-EMG-ET Commit Readiness
NeuroCommitSSM：基于决策的共享自主性框架用于通过EEG-EMG-ET承诺准备性进行安全辅助操作
Abstract
We present NeuroCommitSSM, a decision-centric framework that models when to execute, not just what to do, for safe commit-to-execute control in assistive robotic manipulation. NeuroCommitSSM predicts a continuous commit-readiness score c_t in [0,1] from synchronized electroencephalography (EEG), electromyography (EMG), and eye-tracking (ET), and converts it into discrete commit events through dwell and hysteresis filtering. A three-state finite-state supervisor, HOLD-ASSIST-COMMIT (HAC), gates execution by requiring both a sustained commit-readiness signal from the neural model and real-time perception and robot-state feasibility, including target visibility, inverse kinematics solvability, and collision-free planning, before initiating motion. We evaluate the framework on N=32 subjects performing five activities of daily living (ADL) tasks aligned with the International Classification of Functioning, Disability and Health (ICF), using leave-one-subject-out (LOSO) cross-validation and seven sensor-dropout scenarios (S0-S6). NeuroCommitSSM achieves 0.950 action-balanced accuracy with 0.75 false commit events per 1000 REST windows (FP/1k REST), and maintains low false commits and stable state transitions under sensor loss. For example, in the EEG-only condition, it achieves 0.785 balanced accuracy and 0.29 FP/1k REST, whereas the Temporal Convolutional Network baseline produces 99.95 FP/1k REST under the same condition. Hardware-in-the-loop (HIL) validation on a Kinova Gen3 arm shows that feasibility-checked execution reduces false starts and decision instability without sacrificing task success. Supplementary materials, including code, datasets, videos, and additional analyses, are available at https://madibabaiasl.github.io/NeuroCommitSSM/.
Chinese Translation
我们提出了NeuroCommitSSM，这是一种基于决策的框架，旨在建模何时执行，而不仅仅是做什么，以实现辅助机器人操作中的安全执行控制。NeuroCommitSSM从同步的脑电图（EEG）、肌电图（EMG）和眼动追踪（ET）中预测一个连续的承诺准备性评分c_t，范围在[0,1]之间，并通过停留和滞后过滤将其转换为离散的承诺事件。一个三状态有限状态监督器HOLD-ASSIST-COMMIT（HAC）通过要求神经模型提供持续的承诺准备信号以及实时感知和机器人状态的可行性（包括目标可见性、逆运动学可解性和无碰撞规划）来控制执行，确保在启动运动之前的安全性。我们在N=32名受试者上评估了该框架，受试者执行五项与国际功能、残疾和健康分类（ICF）相关的日常生活活动（ADL）任务，采用留一法交叉验证（LOSO）和七种传感器缺失场景（S0-S6）。NeuroCommitSSM在每1000个REST窗口中实现了0.950的动作平衡准确率和0.75的假承诺事件（FP/1k REST），并在传感器丢失情况下保持低假承诺和稳定的状态转换。例如，在仅使用EEG的条件下，其平衡准确率为0.785，FP/1k REST为0.29，而在相同条件下，时间卷积网络基线产生了99.95 FP/1k REST。在Kinova Gen3机械臂上的硬件在环（HIL）验证显示，经过可行性检查的执行减少了错误启动和决策不稳定性，而不牺牲任务成功率。补充材料，包括代码、数据集、视频和额外分析，均可在https://madibabaiasl.github.io/NeuroCommitSSM/获取。
cs.RO / 3 / 2607.15422
Robust Silicone Pour Casting and Sensor Embedding Procedures for Soft Robotic Actuators
用于软体机器人执行器的稳健硅胶浇铸和传感器嵌入程序
Abstract
Soft robots are well-suited for applications such as rehabilitation and surgery that require adaptable and safe interaction with their environment. However, the challenges of reproducible and scalable fabrication of soft robots limit their real-world deployment. Various fabrication methods have been introduced, but many are labor-intensive and prone to human error. Therefore, traditional two-part pour casting remains an attractive option. This paper presents procedures for robust, repeatable, and scalable fabrication of soft pneumatic actuators using two-part pour casting. The presented methods prevent internal cavity clogging and ensure air-tight sealing. Additionally, a robust sensor embedding procedure for thin-film flex sensors is presented, which allows for accurate and repeatable data acquisition. Finite Element Modeling (FEM) of the soft actuator is performed to analyze stress and deformation from internal pressure loadings. Pneumatic actuation experiments with PID pressure control are performed. Automated image processing is used to calibrate the embedded flex sensor to bending angle measurements. Staircase and sinusoidal profile actuation experiments validate the performance of the fabricated actuator. Angle response experiments for the staircase input show repeatable performance, and the sinusoidal input shows a small amount of hysteresis consistent with viscoelastic response to pneumatic actuation of soft actuators. Simulated and real-world bending angles show comparable response. These methods provide a repeatable and robust fabrication procedure, validated across two operators and 24 successful fabrications, along with benchmark simulations and experimental testing. These benchmarks will enable more widespread adoption of soft robotics.
Chinese Translation
软体机器人非常适合于需要与环境进行适应性和安全互动的应用，如康复和手术。然而，软体机器人的可重复和可扩展制造挑战限制了其在现实世界中的应用。虽然已经提出了多种制造方法，但许多方法劳动密集且容易出现人为错误。因此，传统的双组分浇铸仍然是一个有吸引力的选择。本文提出了一种使用双组分浇铸的稳健、可重复和可扩展的软气动执行器制造程序。所提出的方法防止了内部腔体堵塞，并确保气密密封。此外，本文还提出了一种用于薄膜弯曲传感器的稳健传感器嵌入程序，能够实现准确和可重复的数据采集。对软执行器进行了有限元建模（Finite Element Modeling, FEM），以分析内部压力载荷下的应力和变形。进行了带有PID压力控制的气动驱动实验。使用自动化图像处理对嵌入的弯曲传感器进行标定，以便与弯曲角度测量相匹配。阶梯和正弦波形驱动实验验证了制造的执行器的性能。阶梯输入的角度响应实验显示出可重复的性能，而正弦输入则显示出与软执行器气动驱动的粘弹性响应一致的小量滞后。模拟和实际弯曲角度显示出相似的响应。这些方法提供了一种可重复和稳健的制造程序，经过两个操作员和24次成功制造的验证，以及基准模拟和实验测试。这些基准将促进软体机器人的更广泛应用。
cs.RO / 4 / 2607.15448
VTAP Gripper: Synergizing Fingertip Sensing and a Visuo-Tactile Active Palm for Dexterous In-Hand Manipulation
VTAP抓手：协同指尖感知与视觉触觉主动掌实现灵巧的手内操作
Abstract
This paper presents a tactile-reactive gripper that integrates a Visuo-Tactile Active Palm (VTAP) and compliant, reconfigurable fingers equipped with tactile array sensors. The design exploits structured finger-palm synergy and multi-modal perception to achieve both robust grasping and fine manipulation. The actuated bi-modal palm seamlessly combines long-range visual localization with contact-rich tactile feedback, substantially extending the system's manipulation capability. To bridge the embodiment gap between human hand motion and the heterogeneous three-finger structure, we further propose a staged, gesture-conditioned retargeting framework for dexterous teleoperation. Extensive experiments validate the system across a range of challenging tasks: reactive grasping of YCB and fragile objects, in-hand syringe reorientation and plunger actuation, singulation of clustered objects down to 3 mm in diameter, and vision-tactile peg-in-hole insertion. Results demonstrate that high manipulation performance can be achieved through coordinated finger-palm interaction and multi-modal sensing, without resorting to high degrees of freedom anthropomorphic designs. The VTAP gripper and its retargeting framework offer a practical reference architecture for dexterous gripper design, manipulation, and contact-rich data collection in support of learning-based approaches. Project webpage: https://yuhochau.github.io/vtap/.
Chinese Translation
本文提出了一种触觉反应抓手，该抓手集成了视觉触觉主动掌（VTAP）和配备触觉阵列传感器的柔性可重构手指。该设计利用结构化的指掌协同和多模态感知，实现了稳健的抓取和精细的操作。驱动的双模态掌将远程视觉定位与丰富的接触触觉反馈无缝结合，显著扩展了系统的操作能力。为了弥合人手运动与异构三指结构之间的体现差距，我们进一步提出了一种分阶段的、基于手势的重定向框架，用于灵巧的远程操作。大量实验验证了该系统在一系列挑战性任务中的表现：对YCB和易碎物体的反应性抓取、手内注射器重新定向和活塞驱动、直径缩小到3毫米的聚集物体的单独处理，以及视觉触觉的插销入孔操作。结果表明，通过协调的指掌互动和多模态感知，可以在不依赖高自由度类人设计的情况下实现高操作性能。VTAP抓手及其重定向框架为灵巧抓手设计、操作和支持基于学习的方法进行接触丰富数据收集提供了实用的参考架构。项目网页：https://yuhochau.github.io/vtap/
cs.RO / 5 / 2607.15483
Risk-Aware Preference Learning for Stochastic Outcomes
针对随机结果的风险意识偏好学习
Abstract
Learning reward functions from human preferences is a widely used approach for aligning robot behavior with user expectations in human-robot interaction. Most existing approaches assume that humans evaluate uncertain outcomes using expected utility (EU), aggregating outcome utilities linearly with their probabilities. However, behavioral evidence shows that humans are systematically risk-sensitive, overweighting rare negative events and exhibiting loss aversion. We study the consequences of this mismatch in social robot navigation, where safety-critical outcomes (e.g., collisions) are rare but highly consequential. We compare EU with Cumulative Prospect Theory (CPT), a nonlinear model of human decision-making, within a Bradley-Terry preference learning framework. Our preliminary experiments show that when preferences are generated by risk-sensitive users, CPT-based learners recover reward functions with substantially lower regret compared to EU-based learners. Our results highlight the importance of modeling human risk sensitivity when learning rewards from preferences over stochastic robot outcomes.
Chinese Translation
从人类偏好中学习奖励函数是将机器人行为与用户期望对齐的广泛应用方法，尤其在人与机器人交互中。现有大多数方法假设人类使用期望效用（Expected Utility, EU）来评估不确定结果，即以线性方式将结果效用与其概率进行聚合。然而，行为证据表明，人类在评估时对风险敏感，倾向于高估稀有负面事件，并表现出损失厌恶。我们研究了这种不匹配在社交机器人导航中的后果，其中安全关键的结果（例如碰撞）虽然罕见但后果严重。我们在Bradley-Terry偏好学习框架内比较了EU与累积前景理论（Cumulative Prospect Theory, CPT），后者是一个非线性的人类决策模型。我们的初步实验表明，当偏好由风险敏感的用户生成时，基于CPT的学习者恢复的奖励函数的遗憾值显著低于基于EU的学习者。我们的结果强调了在从随机机器人结果的偏好中学习奖励时，建模人类风险敏感性的重要性。
cs.RO / 6 / 2607.15487
Environment Design for Reliable Shared Autonomy with Probabilistic Guarantees
具有概率保证的可靠共享自主环境设计
Abstract
Shared autonomy enables humans and robots to collaboratively perform tasks by combining human input with autonomous assistance. Most prior work focuses on improving intent inference under a fixed environment, overlooking how workspace design itself affects inference difficulty. We observe that the physical arrangement of objects directly influences the separability of candidate goals under noisy user inputs. We formulate workspace design as an optimization problem and derive a probabilistic correctness guarantee under a bounded noise model. Through simulation experiments across multiple tabletop scenarios, we show that optimized layouts improve goal inference reliability and reduce ambiguity compared to baseline arrangements. We further demonstrate a real-world shared autonomy system that integrates the proposed inference framework. This highlights the role of environment design as a complementary axis for improving shared autonomy systems.
Chinese Translation
共享自主使人类和机器人能够通过结合人类输入与自主辅助来协同执行任务。大多数先前的研究集中在固定环境下改善意图推断，而忽视了工作空间设计本身如何影响推断难度。我们观察到，物体的物理排列直接影响在噪声用户输入下候选目标的可分离性。我们将工作空间设计形式化为一个优化问题，并在有限噪声模型下推导出概率正确性保证。通过在多个桌面场景中的仿真实验，我们展示了优化布局相比于基线排列提高了目标推断的可靠性并减少了歧义。我们进一步展示了一个集成了所提推断框架的现实世界共享自主系统。这突显了环境设计作为改善共享自主系统的补充轴的重要性。
cs.RO / 7 / 2607.15508
Multi-Objective Kinodynamic Motion Planning with Asymptotic Pareto Optimality
具有渐进帕累托最优性的多目标运动规划
Abstract
In this paper, we address the challenge of multi-objective motion planning for systems under kinodynamic constraints. We consider three problem classes: (i) lexicographic optimization, in which objectives are minimized according to a strict priority ordering, (ii) constrained optimization, in which a primary objective is minimized subject to bounds on the remaining costs, and (iii) Pareto front optimization, in which the goal is to approximate the full set of optimal trade-offs among competing objectives. We first show that established cost scalarization methods for multi-objective problems cannot be extended to continuous-domain systems with correctness guarantees. Then, we propose a unified algorithmic framework built upon the Stable Sparse-RRT (SST) algorithm, in which the single representative maintained at each witness neighborhood is replaced by a representative set of locally Pareto-optimal nodes. This structure gives rise to three distinct algorithms: lexSST for lexicographic minimization, coSST for constrained optimization, and poSST for Pareto-front approximation. We provide theoretical guarantees for the completeness and optimality of our algorithms and demonstrate their effectiveness through extensive empirical evaluations.
Chinese Translation
本文针对在运动动力学约束下的多目标运动规划挑战进行探讨。我们考虑三类问题： (i) 字典序优化，其中目标根据严格的优先顺序进行最小化； (ii) 约束优化，其中主要目标在剩余成本的限制下进行最小化； (iii) 帕累托前沿优化，其中目标是近似竞争目标之间的最优权衡的完整集合。我们首先展示了现有的多目标问题成本标量化方法无法扩展到具有正确性保证的连续域系统。然后，我们提出了一个基于稳定稀疏随机树（Stable Sparse-RRT, SST）算法的统一算法框架，其中在每个见证邻域中维护的单一代表被一组局部帕累托最优节点所替代。该结构衍生出三种不同的算法：用于字典序最小化的lexSST，用于约束优化的coSST，以及用于帕累托前沿近似的poSST。我们为这些算法的完整性和最优性提供了理论保证，并通过广泛的实证评估展示了它们的有效性。
cs.RO / 8 / 2607.15579
PACE: Persona Adaptation through Conversational Elicitation in Human-Robot Interaction
PACE：通过对话引导实现人机交互中的角色适应
Abstract
Equipping humanoid robots with coherent and adaptable personas is crucial for fostering natural, engaging, and trustworthy human-robot interaction (HRI). However, existing approaches often rely on static, hard-coded identities that lack the flexibility to adapt to individual user contexts. In this paper, we present PACE (Persona Adaptation through Conversational Elicitation), a novel framework for the interactive generation and deployment of structured personas on the Ameca humanoid robot. Our system introduces an Interactive Persona Elicitation Pipeline, enabling the robot to dynamically synthesize a tailored, psychologically grounded identity through user Q&A. This elicitation process feeds into a persona prompt compilation phase, generating a structured persona prompt built upon multi-perspective dimensions. We detail the Embodied System Integration required to translate this structured specification into expressive, multimodal humanoid behaviors. Through a comprehensive empirical HRI evaluation, we assess the impact of dynamically generated personas on user trust, perceived anthropomorphism, persona consistency, personal relevance, and interaction quality compared to a generic baseline. These contributions establish a scalable pathway for deploying personalized, interactive, and reliable identities in embodied humanoid assistants. Video demo is available at: https://lipzh5.github.io/PACE/
Chinese Translation
为类人机器人赋予连贯且可适应的角色对于促进自然、引人入胜和可信赖的人机交互（HRI）至关重要。然而，现有的方法往往依赖于静态的、硬编码的身份，缺乏适应个体用户上下文的灵活性。本文提出了PACE（通过对话引导实现角色适应），这是一个用于在Ameca类人机器人上交互生成和部署结构化角色的新框架。我们的系统引入了一个交互式角色引导管道，使机器人能够通过用户问答动态合成量身定制的、心理学基础的身份。该引导过程进入角色提示编译阶段，生成基于多角度维度的结构化角色提示。我们详细描述了将这一结构化规范转化为富有表现力的多模态类人行为所需的具身系统集成。通过全面的实证HRI评估，我们评估了动态生成的角色对用户信任、感知拟人化、角色一致性、个人相关性和交互质量的影响，并与通用基线进行了比较。这些贡献为在具身类人助手中部署个性化、互动和可靠的身份建立了可扩展的途径。视频演示可在以下链接查看：https://lipzh5.github.io/PACE/
cs.RO / 9 / 2607.15582
A Model-Based Decoupling Strategy for Proprioception and Contact Sensing in an Architected Soft Manipulator
基于模型的自我感知与接触感知解耦策略在架构软操纵器中的应用
Abstract
Soft continuum robots require embedded sensing for proprioception and contact detection, yet integrating sensors into sparse, highly deformable architected structures remains challenging. We present a model-based strategy that decouples proprioceptive and contact signals from a common set of fluidic pressure sensors embedded in a soft architected segment. Each segment of the Innervated Trimmed Helicoid (ITH) contains six air channels routed in a localized zigzag pattern along the circumference. With only three principal kinematic degrees of freedom (axial compression, bending in x, bending in y), the six pressure readings form an overdetermined system. A piecewise constant curvature model maps pressures to shape, and Huber regression identifies outlier channels whose residuals indicate external contact. On a single ITH segment, this approach achieves proprioceptive shape estimation with a relative bending error of 0.11 +/- 0.02 and a contact detection rate of 97% across 178 trials. We integrate eight ITH segments into Air-Helix, a tendon-driven soft continuum manipulator, and present exploratory whole-arm demonstrations that include tactile teaching by demonstration, admittance-controlled force regulation, and tactile object reconstruction. The results suggest that localized fluidic innervation combined with model-based redundancy resolution is a practical path toward concurrent proprioception and contact sensing in architected soft robots.
Chinese Translation
软连续机器人需要嵌入式传感器以实现自我感知和接触检测，但将传感器集成到稀疏且高度可变形的架构结构中仍然具有挑战性。我们提出了一种基于模型的策略，该策略将自我感知和接触信号从嵌入在软架构段中的一组流体压力传感器中解耦。每个神经调节修剪螺旋体（Innervated Trimmed Helicoid, ITH）段包含六个沿周长以局部之字形模式布置的气道。仅凭三个主要运动自由度（轴向压缩、x方向弯曲、y方向弯曲），这六个压力读数形成一个超定系统。分段常数曲率模型将压力映射到形状，Huber回归识别出其残差指示外部接触的异常通道。在单个ITH段上，该方法实现了自我感知形状估计，相对弯曲误差为0.11 +/- 0.02，接触检测率在178次试验中达到97%。我们将八个ITH段集成到气螺旋（Air-Helix）中，这是一种由腱驱动的软连续操纵器，并展示了包括通过示范进行触觉教学、基于导纳的力调节和触觉物体重建在内的全臂探索性演示。结果表明，局部流体神经调节结合基于模型的冗余解决方案是实现架构软机器人中自我感知与接触感知并行的实用途径。
cs.RO / 10 / 2607.15589
MemoGuard: An Adaptive Runtime for Guarding Against Memory Traps in Communication-Limited Robot Navigation
MemoGuard：一种自适应运行时，用于防范通信受限机器人导航中的内存陷阱
Abstract
Communication-limited robots in mission-critical scenarios such as disaster inspection and search-and-rescue must make reliable onboard decisions without access to remote operators or high-capacity reasoning services. Episodic memory reuse is an attractive low-cost fallback, but retrieval similarity does not guarantee execution validity, i.e., a retrieved action may match the current context yet be unsafe due to changed topology, insufficient battery margin, or unreliable prior outcomes. We call such high-similarity but execution-invalid episodes memory traps. This creates a safety-efficiency design space where similarity only reuse minimizes fallback cost but can be unsafe, while always invoking local reasoning improves safety at high computational and energy cost. This paper presents MemoGuard, a lightweight adaptive runtime that validates episodic memories against topology, resource, and outcome contracts before reuse, invoking fallback only when validation fails. In a graph-based corridor-inspection simulator, MemoGuard reduces battery safety violations by 76.6% over similarity-only top-1 reuse while reducing fallback calls by 21.4% over always reasoning. On an NVIDIA Jetson AGX Xavier with local llama3.2:3b fallback reasoning, this corresponds to 3.67 s and 36.97 J of avoided fallback-reasoning overhead per trial. We open-source MemoGuard at https://github.com/hetheiin/memoguard.
Chinese Translation
在灾难检查和搜索救援等关键任务场景中，通信受限的机器人必须在没有远程操作员或高容量推理服务的情况下做出可靠的 onboard 决策。情节记忆重用是一种有吸引力的低成本后备方案，但检索相似性并不能保证执行的有效性，即检索到的动作可能与当前上下文匹配，但由于拓扑变化、电池余量不足或不可靠的先前结果而变得不安全。我们称这种高相似性但执行无效的情节为内存陷阱。这创造了一个安全性与效率的设计空间，其中仅依赖相似性重用可以最小化后备成本，但可能不安全，而始终调用本地推理则在高计算和能量成本下提高安全性。本文提出了 MemoGuard，一种轻量级自适应运行时，它在重用之前验证情节记忆是否符合拓扑、资源和结果合同，仅在验证失败时调用后备。在基于图的走廊检查模拟器中，MemoGuard 将电池安全违规率降低了 76.6%，相比于仅依赖相似性进行的 top-1 重用，同时将后备调用减少了 21.4%，相比于始终进行推理。在配备本地 llama3.2:3b 后备推理的 NVIDIA Jetson AGX Xavier 上，这相当于每次试验避免了 3.67 秒和 36.97 焦耳的后备推理开销。我们在 https://github.com/hetheiin/memoguard 上开源了 MemoGuard。
cs.RO / 11 / 2607.15620
AEGIS: Assay-Aware Protocol Validation and Runtime Monitoring for Open-Source Liquid Handling Robots
AEGIS：针对开源液体处理机器人进行实验检测感知的协议验证与运行时监控
Abstract
Self-driving laboratories increasingly rely on low-cost liquid handlers such as the Opentrons OT-2, which ship without the pressure-based aspiration monitoring of Hamilton or Tecan systems and are typically run open-loop. Two failure modes go undetected: protocols that are syntactically valid but violate assay-specific invariants (e.g., tip reuse between a PCR template and a no-template control), and physical execution failures (partial dispense, air bubbles, missing tips) at runtime. We present AEGIS, a two-layer guardian for both. Layer 1 pairs a curated machine-readable assay rule database with an LLM that reasons over OT-2 Python code, reaching an adjusted F1 of 0.97 on a 24-protocol benchmark across five assay families and beating rules-only and LLM-only ablations across five backends; a free open-weight model ties the best proprietary one, so no paid API is required. Layer 2 fits a PCA world model to YOLO-cropped four-frame pipette trajectories; under a leakage-free leave-one-plate-out evaluation it reaches average precision 0.89 and operating-point F1 0.71 (AUROC 0.80), a deployment-faithful number that matches the live demonstration, and we characterize the small-pipette (p20) resolution limit (F1 0.47). A live demonstration on a physical OT-2 (five replicates per condition) catches planted no-tip failures deterministically and partial dispense on coloured dyes, with an always-VLM self-vote gate lifting partial-dispense recall to 5/5; transparent water is a principled limit of any front-view-only monitor, which AEGIS surfaces as low-confidence VLM reasoning rather than a wrong verdict. Cascade triage holds VLM cost near $1.63 per plate versus $10.33 for an always-VLM baseline. AEGIS is open source and, to our knowledge, the first system to unify pre-flight assay-aware validation with runtime visual monitoring for an open-source liquid handler.
Chinese Translation
自驾实验室越来越依赖于低成本的液体处理器，如 Opentrons OT-2，这些设备在出厂时没有 Hamilton 或 Tecan 系统的基于压力的吸取监控，通常以开放回路运行。有两种故障模式未被检测到：一是语法上有效但违反特定实验检测不变式的协议（例如，在 PCR 模板和无模板对照之间重复使用吸头），二是在运行时发生的物理执行故障（部分分配、气泡、缺失吸头）。我们提出了 AEGIS，一个针对这两种情况的双层守护系统。第一层将一个经过筛选的机器可读实验规则数据库与一个能够推理 OT-2 Python 代码的大型语言模型（LLM）相结合，在五个实验类别的 24 个协议基准测试中达到了 0.97 的调整 F1 值，超越了仅使用规则和仅使用 LLM 的消融实验，适用于五个后端；一个免费的开放权重模型与最佳的专有模型相当，因此不需要付费 API。第二层将主成分分析（PCA）世界模型应用于 YOLO 裁剪的四帧移液轨迹；在无泄漏的留一板评估下，平均精度达到了 0.89，操作点 F1 值为 0.71（AUROC 0.80），这一部署忠实的数字与现场演示相匹配，同时我们还表征了小吸头（p20）的分辨率极限（F1 0.47）。在物理 OT-2 上的现场演示（每种条件五次重复）能够确定性地捕捉到植入的无吸头故障和在有色染料上的部分分配，始终-VLM 自投票门将部分分配的召回率提升至 5/5；透明水是任何仅前视监控的原则性限制，AEGIS 将其呈现为低置信度的 VLM 推理，而非错误的裁决。级联分流使 VLM 成本保持在每板约 1.63 美元，而始终-VLM 基线为 10.33 美元。AEGIS 是开源的，并且据我们所知，是第一个将飞行前实验检测感知验证与开源液体处理器的运行时视觉监控相结合的系统。
cs.RO / 12 / 2607.15621
Think at 5 Hz, Act at 20 Hz: Asynchronous Fast-Slow Vision-Language-Action Inference for Closed-Loop Driving
以5 Hz思考，以20 Hz行动：闭环驾驶中的异步快慢视觉-语言-动作推断
Abstract
Large language models bring instruction following and scene reasoning to end-to-end driving, but their inference latency collides with the control rate a vehicle requires. Existing closed-loop agents hide this gap by invoking the model on alternate simulation ticks and replaying the previous command in between, so half of all control outputs ignore the newest observations. We present a fast-slow architecture that removes this compromise. A frozen 7B vision-language backbone acts as the slow system, digesting navigation instructions and visual history at low frequency while exposing its per-layer key-value cache as a standing representation of the scene. A lightweight action expert acts as the fast system, attending to this cache and to the current camera frame at every simulation tick to regress waypoints in a single forward pass. Since the cache lags behind the world at deployment, we train the expert under randomized staleness, aligning training with asynchronous execution. On LangAuto-Short routes in CARLA, our system produces fresh control at every 50 ms simulation tick and lifts route completion from 37.0 to 94.0 over the frame-skipping baseline. A frame-skip ablation with the same expert separates the two factors at work: the expert raises the driving score on its own, while per-tick freshness raises completion from 82.1 to 94.0 and cuts red-light violations by a third. Trained on a single town, the expert transfers zero-shot to two unseen towns, holding 84-94% route completion where the baseline reaches 31-41%. It reduces open-loop waypoint error by nearly a factor of four compared to the backbone's own action head, at a per-tick model cost of 32 ms that is independent of history length on a single consumer GPU.
Chinese Translation
大型语言模型将指令跟随和场景推理引入端到端驾驶，但它们的推断延迟与车辆所需的控制频率相冲突。现有的闭环代理通过在交替的仿真时刻调用模型，并在两者之间重放先前的命令，来掩盖这一差距，因此一半的控制输出忽略了最新的观察结果。我们提出了一种快慢架构，消除了这种妥协。一个冻结的7B视觉-语言主干作为慢系统，以低频率消化导航指令和视觉历史，同时将其每层的键值缓存作为场景的静态表示。一个轻量级的动作专家作为快系统，在每个仿真时刻关注该缓存和当前的相机帧，以单次前向传递回归航点。由于缓存在部署时滞后于现实世界，我们在随机陈旧性下训练专家，使训练与异步执行对齐。在CARLA的LangAuto-Short路线中，我们的系统在每50毫秒的仿真时刻产生新鲜的控制，并将路线完成率从37.0提升至94.0，超越了跳帧基线。与相同专家的跳帧消融实验分离了两种作用因素：专家单独提高了驾驶评分，而每个时刻的新鲜度将完成率从82.1提升至94.0，并减少了三分之一的红灯违规。专家在一个城镇上训练后，零样本迁移到两个未见过的城镇，保持84-94%的路线完成率，而基线仅达到31-41%。与主干的动作头相比，它将开放循环航点误差降低了近四倍，在单个消费级GPU上，每个时刻的模型成本为32毫秒，与历史长度无关。
cs.RO / 13 / 2607.15633
Scalable Open-Source Visuotactile Sensor for 6-Axis Contact Wrench Estimation in Tensegrity Robots
可扩展的开源视觉触觉传感器用于张力结构机器人中的六轴扭矩估计
Abstract
This paper presents a scalable, open-source visuotactile sensing system for tensegrity robots that enables six-axis wrench estimation and contact detection. The proposed endcap sensor integrates an elastomeric shell, a 3D-printed thermoplastic polyurethane (TPU) interface, and a rigid base housing an embedded camera and LED illumination ring. A novel gyroid-infill bonding technique is introduced to form a durable elastomer-TPU interface without adhesives, yielding a lightweight and modular design compatible with large-scale tensegrity structures. A tactile-to-wrench neural network maps shear vector fields to six-dimensional force and torque measurements. Experimental results demonstrate accurate and stable wrench estimation with a mean squared error (MSE) of 0.1531 on static validation data and out-of-domain generalization under dynamic motion. Furthermore, full-system integration on a 12 kg tensegrity robot confirms the sensor's ability to reliably identify ground contacts. The system substantially improves the practicality of tactile feedback for tensegrity robots, offering a low-cost, reproducible, and physically interpretable pathway toward contact-aware proprioception and state estimation. Open source files are available at \href{https://github.com/Jonathan-Twz/tensegrity-gelfoot}{github.com/Jonathan-Twz/tensegrity-gelfoot}
Chinese Translation
本文提出了一种可扩展的开源视觉触觉传感系统，适用于张力结构机器人，实现六轴扭矩估计和接触检测。所提出的端盖传感器集成了弹性外壳、3D打印热塑性聚氨酯（TPU）界面和一个包含嵌入式摄像头及LED照明环的刚性底座。引入了一种新颖的 gyroid-infill 结合技术，以无粘合剂的方式形成耐用的弹性体-TPU 接口，从而实现轻量化和模块化设计，兼容大规模张力结构。触觉到扭矩的神经网络将剪切矢量场映射到六维力和扭矩测量。实验结果表明，在静态验证数据上，扭矩估计的均方误差（MSE）为0.1531，且在动态运动下具有跨域泛化能力。此外，在一台12公斤的张力结构机器人上进行的全系统集成验证了传感器可靠识别地面接触的能力。该系统显著提高了张力结构机器人的触觉反馈实用性，提供了一种低成本、可重复且具有物理可解释性的途径，朝向接触感知的本体感知和状态估计。开源文件可在 https://github.com/Jonathan-Twz/tensegrity-gelfoot 获取。
cs.RO / 14 / 2607.15641
IMBench: A Benchmark for Intuitive Robotic Manipulation
IMBench：直观机器人操控的基准测试
Abstract
Humans combine reasoning and motor control to solve complex manipulation tasks under diverse constraints. They build an understanding of the physical world that helps them convert reasoning into actions and quickly adapt to new scenes, tasks, and rules. We refer to this capability as intuitive manipulation. Existing benchmarks fail to capture this integration: they evaluate physical reasoning in isolation from execution, or measure policy performance without requiring explicit reasoning. We introduce IMBENCH, a benchmark designed to evaluate intuitive manipulation as an integrated capability spanning perception, physical reasoning, action generation, and iterative execution. Our tasks require models to infer task-relevant physical structure and generate feasible action sequences under explicit constraints, including contact-rich manipulation, tool use, and multi-stage dependencies. We introduce a benchmark of 35 tasks, 14K filtered trajectories, and scalable tools for generating diverse scenarios. Experiments reveal a consistent gap: vision language models show partial physical reasoning ability but fail to produce executable plans, while state-of-the-art vision-language-action models struggle to satisfy task constraints and generalize across scenarios. These results identify intuitive manipulation as a missing axis in current foundation models and generalist robot policies, and position IMBENCH as a step toward evaluating and enabling more integrated, adaptive physical intelligence.
Chinese Translation
人类结合推理与运动控制，以在多样化的约束下解决复杂的操控任务。他们建立对物理世界的理解，这帮助他们将推理转化为行动，并迅速适应新的场景、任务和规则。我们将这种能力称为直观操控。现有基准测试未能捕捉这种整合：它们要么孤立地评估物理推理与执行，要么在不要求明确推理的情况下测量策略性能。我们引入IMBENCH，这是一个旨在评估直观操控作为一种跨越感知、物理推理、动作生成和迭代执行的整合能力的基准测试。我们的任务要求模型推断与任务相关的物理结构，并在明确约束下生成可行的动作序列，包括接触丰富的操控、工具使用和多阶段依赖关系。我们引入了一个包含35个任务、14K个过滤轨迹和可扩展工具以生成多样化场景的基准测试。实验结果揭示了一个持续的差距：视觉语言模型显示出部分物理推理能力，但未能生成可执行的计划，而最先进的视觉-语言-动作模型在满足任务约束和跨场景泛化方面存在困难。这些结果将直观操控识别为当前基础模型和通用机器人策略中缺失的一个维度，并将IMBENCH定位为评估和促进更整合、适应性强的物理智能的一个步骤。
cs.RO / 15 / 2607.15642
Difference-Based Relational Learning for Zero-Shot Object-Goal Visual Navigation With Direct Sim-to-Real Transfer
基于差异的关系学习用于零样本目标视觉导航的直接模拟到现实转移
Abstract
End-to-end deep reinforcement learning (DRL) for zero-shot object-goal visual navigation remains challenged by the sim-to-real gap, particularly variations in object appearance and restricted camera field-of-view (FoV). This letter proposes a Temporal Difference-Relational Network (T-DRN) for robust zero-shot sim-to-real transfer. T-DRN combines a Siamese difference-based feature extractor, which computes relational difference between the target and observed objects to produce domain-independent representations, with a dual-frame temporal buffer that preserves short-term object continuity under narrow FoV. Extensive experiments in AI2-THOR demonstrate that T-DRN improves zero-shot generalization in terms of success rates over strong baselines. Furthermore, T-DRN is systematically validated on a physical wheeled robot, demonstrating robust performance under real sensing and actuation constraints and supporting the feasibility of direct sim-to-real transfer.
Chinese Translation
端到端深度强化学习（DRL）在零样本目标视觉导航中仍面临模拟到现实的差距挑战，特别是在物体外观变化和受限的相机视场（FoV）方面。本文提出了一种时间差异关系网络（Temporal Difference-Relational Network, T-DRN），用于实现稳健的零样本模拟到现实转移。T-DRN结合了一个基于差异的Siamese特征提取器，该提取器计算目标与观察到的物体之间的关系差异，以生成领域无关的表示，以及一个双帧时间缓冲区，该缓冲区在狭窄的FoV下保持短期物体连续性。在AI2-THOR中的大量实验表明，T-DRN在成功率方面改善了零样本泛化，超越了强基线。此外，T-DRN在物理轮式机器人上进行了系统验证，展示了在真实传感和执行约束下的稳健性能，支持了直接模拟到现实转移的可行性。
cs.RO / 16 / 2607.15656
Learning a System-Level Surrogate for Hydraulic Excavators: A Simulation-to-Real LSTM Approach
为液压挖掘机学习系统级代理模型：一种从仿真到现实的长短期记忆网络方法
Abstract
Developing autonomous hydraulic excavators is constrained by limited access to physical machines and the high cost of real-world experimentation. This paper proposes a simulation-to-real framework for learning a system-level digital surrogate using Long Short-Term Memory (LSTM) networks. Instead of modeling internal dynamics, the excavator is treated as an input-output operator, and the surrogate is trained to reproduce its closed-loop behavior under identical control inputs. The approach is first validated in a MuJoCo simulation environment and then transferred to a real excavator. To address measurement inconsistencies in real-world data, a consistency-aware state estimation method based on adaptive Kalman filtering is introduced. Experimental results demonstrate that the learned surrogate achieves high fidelity in both angular velocity and long-horizon trajectory reproduction under closed-loop autoregressive evaluation. These results confirm that the proposed model can serve as a drop-in surrogate for both simulation and physical systems, enabling scalable and efficient development of excavation automation algorithms.
Chinese Translation
开发自主液压挖掘机受到物理机器获取有限和现实世界实验高成本的限制。本文提出了一种从仿真到现实的框架，用于利用长短期记忆网络（LSTM）学习系统级数字代理模型。该方法不再建模内部动态，而是将挖掘机视为输入-输出操作符，代理模型被训练以在相同控制输入下重现其闭环行为。该方法首先在MuJoCo仿真环境中进行验证，然后转移到真实挖掘机上。为了解决现实世界数据中的测量不一致性，提出了一种基于自适应卡尔曼滤波的状态估计方法。实验结果表明，所学习的代理模型在闭环自回归评估下，在角速度和长时间轨迹重现方面都达到了高保真度。这些结果确认了所提出的模型可以作为仿真和物理系统的直接替代品，从而促进挖掘自动化算法的可扩展和高效开发。
cs.RO / 17 / 2607.15659
Continuously Stable Structure through Plastic Deformation
通过塑性变形实现持续稳定结构
Abstract
Soft robots have seen widespread adoption in interactive tasks due to their inherent compliance and adaptability. However, these advantages often come at the cost of stability, posing challenges in a dynamic environment. This limitation is especially critical in soft grippers, where instability under acceleration or external disturbances can result in grasp failure. In this study, we present a continuously stable structure through plastic deformation (CSSPD), integrated into a soft gripper. By leveraging the mechanism of plastic deformation, the gripper maintains continuous configurations without energy input, while the added stiffness ensures both static and dynamic stability. We introduce a bioinspired paw pad that significantly enhances stability and enables sensing-based rapid object grasping. Then we develop the mathematical model and optimize the kirigami structure of the metal layer. Experimental results show that the gripper can sustain a passive holding force of up to 16 N without energy input, achieving performance comparable to pneumatic actuation at 0.3 MPa. When combined with pneumatic actuation, it remains stable under pulsed accelerations of up to 400 m/s^2. It can also passively perch on tree branches for extended periods without power, demonstrating promise for mobile robotic applications.
Chinese Translation
软体机器人因其固有的柔韧性和适应性在交互任务中得到了广泛应用。然而，这些优势往往以牺牲稳定性为代价，在动态环境中带来了挑战。这一限制在软抓手中尤为关键，因为在加速或外部干扰下的不稳定性可能导致抓取失败。在本研究中，我们提出了一种通过塑性变形（CSSPD）实现的持续稳定结构，并将其集成到软抓手中。通过利用塑性变形机制，抓手能够在不输入能量的情况下维持连续的配置，同时增加的刚度确保了静态和动态的稳定性。我们引入了一种仿生爪垫，显著增强了稳定性并实现基于传感的快速物体抓取。接着，我们开发了数学模型并优化了金属层的切纸结构。实验结果表明，抓手在不输入能量的情况下能够维持高达16 N的被动保持力，性能达到与0.3 MPa气动驱动相当的水平。当与气动驱动结合时，它在高达400 m/s²的脉冲加速度下仍保持稳定。它还能够在树枝上被动栖息较长时间而无需供电，展示了在移动机器人应用中的潜力。
cs.RO / 18 / 2607.15674
A Generative Partially Specified Finite State Machine Approach to Complex Behaviour Planning
一种生成性部分指定有限状态机方法用于复杂行为规划
Abstract
Autonomous robots operating in dynamic environments require behaviour planning systems that combine reactivity, interpretability, and adaptability. While Large Language Models have been successfully integrated with Behaviour Trees for dynamic replanning, Finite State Machines, despite their widespread adoption and computational efficiency, remain unexplored for generative approaches. We propose a Generative Partially Specified Finite State Machine (GPSFSM) neurosymbolic architecture that utilises the symbolic and semantic structure of FSMs, including states and event-triggered transitions, to implement Behaviour Planning. This paper introduces the first GPSFSM framework for robotics, featuring Fabric, an FSM engine that parses, validates, and executes behaviour plans that contain Sequential, Recovery, Parallel-Any, and Parallel-All control structures. We extend the Capabilities2 package in ROS2 with an asynchronous event system for behaviour chaining and runtime parameter injection for configurable execution, addressing the ad-hoc function representations that limit current generative systems. PromptTools provides a unified ROS 2 interface to local and cloud LLMs, with prompt buffering, enabling dynamic asynchronous composition of task and context information. Together, these components enable standardised semantic capability descriptions for robot-agnostic development. Experimental evaluation on navigation tasks demonstrates that our GPSFSM approach achieves consistently higher plan-generation success rates than the state-of-the-art BTGenBot system, particularly excelling in zero-shot scenarios where BTs typically struggle, while maintaining comparable or lower planning latency to frontier LLMs. We also demonstrate that our system can generate complex behaviours. We release an open-source ROS2 stack that makes generative FSM planning practical and reproducible for robotic systems.
Chinese Translation
在动态环境中运行的自主机器人需要结合反应性、可解释性和适应性的行为规划系统。尽管大型语言模型已成功与行为树结合用于动态重新规划，但有限状态机（Finite State Machines, FSM）尽管广泛应用且计算效率高，仍未在生成性方法中得到探索。我们提出了一种生成性部分指定有限状态机（Generative Partially Specified Finite State Machine, GPSFSM）神经符号架构，利用FSM的符号和语义结构，包括状态和事件触发的转移，来实现行为规划。本文介绍了首个针对机器人技术的GPSFSM框架，包含Fabric，一个解析、验证和执行包含顺序、恢复、并行任意和并行全控制结构的行为计划的FSM引擎。我们在ROS2中扩展了Capabilities2包，增加了用于行为链的异步事件系统和可配置执行的运行时参数注入，解决了当前生成系统中限制性的临时功能表示问题。PromptTools提供了一个统一的ROS 2接口，连接本地和云端的大型语言模型（LLMs），并具备提示缓冲功能，支持任务和上下文信息的动态异步组合。这些组件共同实现了机器人无关开发的标准化语义能力描述。在导航任务上的实验评估表明，我们的GPSFSM方法在计划生成成功率上始终高于最先进的BTGenBot系统，特别是在行为树通常表现不佳的零-shot场景中，同时保持与前沿LLMs相当或更低的规划延迟。我们还展示了我们的系统能够生成复杂行为。我们发布了一个开源的ROS2堆栈，使生成性FSM规划在机器人系统中变得实用和可重复。
cs.RO / 19 / 2607.15701
RAVEN: Reinforcement-Adaptive Visibility-Graph Planning for Robust Humanoid Navigation with Collision-Free MPC
RAVEN：用于鲁棒人形导航的强化自适应可视图规划与无碰撞模型预测控制
Abstract
Humanoid navigation in dynamic environments requires long-horizon planning while respecting short-horizon dynamic and safety constraints. Classical visibility-graph planners combined with model predictive control (MPC) can efficiently generate collision-free trajectories, but their performance depends on manually tuned parameters and accurate system modeling. In real robotic systems, control delays, state-estimation noise, and locomotion uncertainties can cause overshoot and constraint violations even when the nominal path is geometrically optimal. We propose RAVEN, a hierarchical reinforcement learning (RL)-MPC framework for robust humanoid navigation. Unlike prior approaches that use learning to tune cost weights or replace planning entirely, RAVEN employs RL to adapt the geometric construction of a visibility-graph planner by modifying obstacle inflation and related graph parameters. By directly reshaping the free-space geometry, the learned planner alters the topology of the global path to compensate for delay and tracking imperfections. A collision-free MPC layer then tracks the planned trajectory while explicitly enforcing velocity bounds and obstacle-avoidance constraints. By training under realistic delays and observation noise, RAVEN learns planning adaptations that improve robustness while retaining explicit long-horizon geometric planning and constrained optimization, in contrast to end-to-end learning approaches. We evaluate RAVEN against a manually tuned visibility-graph MPC baseline and a pure RL navigation policy. Results demonstrate reduced overshoot near obstacles, improved robustness in narrow passages, and more reliable navigation under delay and noise. These findings indicate that reinforcement-adaptive graph construction combined with constrained MPC provides an effective and interpretable alternative to end-to-end learning for robust humanoid navigation.
Chinese Translation
在动态环境中，人形导航需要进行长时间规划，同时遵循短时间的动态和安全约束。经典的可视图规划器结合模型预测控制（MPC）可以有效生成无碰撞的轨迹，但其性能依赖于手动调节的参数和准确的系统建模。在真实的机器人系统中，控制延迟、状态估计噪声和运动不确定性可能导致超调和约束违反，即使在几何上最优的路径下也是如此。我们提出了RAVEN，一个用于鲁棒人形导航的分层强化学习（RL）-MPC框架。与之前使用学习来调节成本权重或完全替代规划的方法不同，RAVEN利用RL通过修改障碍物膨胀和相关图参数来适应可视图规划器的几何构造。通过直接重塑自由空间几何，学习到的规划器改变了全局路径的拓扑，以补偿延迟和跟踪缺陷。然后，一个无碰撞的MPC层跟踪规划的轨迹，同时明确执行速度限制和避障约束。通过在现实的延迟和观测噪声下进行训练，RAVEN学习到的规划适应性提高了鲁棒性，同时保留了明确的长时间几何规划和约束优化，这与端到端学习方法形成对比。我们将RAVEN与手动调节的可视图MPC基线和纯RL导航策略进行了评估。结果表明，在障碍物附近减少了超调，在狭窄通道中提高了鲁棒性，并在延迟和噪声下提供了更可靠的导航。这些发现表明，结合约束MPC的强化自适应图构造为鲁棒人形导航提供了一种有效且可解释的替代方案，优于端到端学习。
cs.RO / 20 / 2607.15708
Implicit Virtual Leader: Decentralized Vision-Only Relative Pose Estimation for Multi-Robot Formations
隐式虚拟领导者：基于视觉的多机器人编队去中心化相对位姿估计
Abstract
Classical leader-follower formation control suffers from single points of failure and error propagation, and relies on absolute localization sensors that are ill-suited for GPS-denied environments. We address these limitations by introducing a fully decentralized, vision-only relative pose estimation framework based on Graph Neural Networks (GNNs). The key idea is the implicit virtual leader (IVL): a non-physical formation reference frame that is not tied to any individual robot but is implicitly learned within the GNN using only monocular images and inter-robot communication. We attach a heteroscedastic GNLL head for aleatoric uncertainty and MC~Dropout for epistemic uncertainty, and conduct a systematic comparison across simulation and real-world test sets. Our framework achieves competitive pose estimation accuracy and generalizes naturally to heterogeneous robot platforms and varying formation sizes.
Chinese Translation
传统的领导-跟随编队控制存在单点故障和误差传播的问题，并依赖于不适合GPS缺失环境的绝对定位传感器。我们通过引入一个完全去中心化的、仅基于视觉的相对位姿估计框架来解决这些限制，该框架基于图神经网络（Graph Neural Networks, GNNs）。其关键思想是隐式虚拟领导者（Implicit Virtual Leader, IVL）：一个非物理的编队参考框架，它不依赖于任何单个机器人，而是通过仅使用单目图像和机器人间通信在GNN中隐式学习。我们为随机不确定性附加了异方差的GNLL头，并为认知不确定性使用了MC Dropout，并在仿真和真实世界测试集上进行了系统比较。我们的框架实现了具有竞争力的位姿估计精度，并自然地推广到异构机器人平台和不同的编队规模。
cs.RO / 21 / 2607.15714
AC-VLA: Robust Out-of-Distribution Action Execution via Compositional Learning
AC-VLA：通过组合学习实现稳健的分布外动作执行
Abstract
Vision-Language-Action (VLA) models excel at end-to-end robotic manipulation but struggle with out-of-distribution (OOD) generalization when familiar sub-tasks are recombined in unseen configurations. We identify two mutually reinforcing failure modes: \emph{trajectory overfitting}, where models overfit to holistic trajectory patterns rather than compositional sub-skill semantics; and \emph{perceptual shortcut}, where action tokens over-rely on wrist-view textures at the expense of global spatial grounding. To address both, we introduce \textbf{AC-VLA}, a plug-and-play Action Compositional learning framework comprising two architecture-agnostic components: \textbf{(i)} a compositional learning module that uses an LLM-driven instruction decomposer and a proprioceptive trajectory aligner to generate dense sub-task supervision, followed by mixed training on complete demonstrations and decomposed data to endow the model with compositional generalization; and \textbf{(ii)} a state-conditioned asymmetric masking strategy that suppresses wrist-view inputs during closed-gripper phases, enforcing global semantic grounding. All components are architectural modification-free and directly integrable into any VLA backbone. Instantiated on $\pi_{0.5}$ and evaluated on LIBERO and LIBERO-OOD benchmarks, AC-VLA achieves a ~28% absolute improvement on compositional OOD tasks while maintaining near-perfect in-distribution performance.
Chinese Translation
视觉-语言-动作（VLA）模型在端到端的机器人操作中表现出色，但在熟悉子任务以未见配置重新组合时，面临分布外（OOD）泛化的挑战。我们识别出两种相互强化的失败模式： extit{轨迹过拟合}，即模型过度拟合整体轨迹模式而非组合子技能语义；以及 extit{感知捷径}，即动作标记过度依赖手腕视角的纹理，而忽视全局空间的基础。为了解决这两个问题，我们引入了 extbf{AC-VLA}，一个即插即用的动作组合学习框架，包含两个与架构无关的组件： extbf{(i)} 一个组合学习模块，使用基于大型语言模型（LLM）的指令分解器和本体轨迹对齐器生成密集的子任务监督，随后在完整演示和分解数据上进行混合训练，以赋予模型组合泛化能力；以及 extbf{(ii)} 一种状态条件的不对称掩蔽策略，在闭合夹持阶段抑制手腕视角输入，从而强制执行全局语义基础。所有组件均无架构修改，且可直接集成到任何VLA主干网络中。在$ extpi_{0.5}$上实例化，并在LIBERO和LIBERO-OOD基准上进行评估，AC-VLA在组合OOD任务上实现了约28%的绝对提升，同时保持近乎完美的分布内性能。
cs.RO / 22 / 2607.15733
A Task-Space Receding Horizon Controller for Fast Collision Avoidance
用于快速避碰的任务空间递归控制器
Abstract
Real-time collision avoidance for robotic manipulators requires fast reactions to unexpected obstacle motion and lookahead to avoid becoming trapped by near-future constraints. Full model predictive control can provide this foresight, but its online cost may grow quickly with horizon length, model fidelity, and the number of active geometric constraints. Conversely, horizon-free reactive methods are computationally efficient but can be short-sighted in dynamic clutter. We present a task-space receding-horizon controller that uses a short contact-consistent rollout to generate a terminal kinematic reference satisfying internal non-penetration constraints, then computes only the first input of a smooth minimum-acceleration transition toward that reference. Starting from a closed-loop inverse-kinematics regulation law, the rollout is performed with an iterative dynamics solver operating on inflated convex robot and obstacle geometries, so that robot-obstacle contacts, dynamic obstacle motion, and self-collisions can shape the terminal reference without requiring full constrained trajectory optimization. We analyze the contact-inactive closed loop and show local exponential task-space regulation under standard regularity assumptions. For contacts activated inside the rollout, we characterize the corresponding discrete updates and bound the effect of moving obstacles on regular operating sets. Simulations on a 40-DOF multi-chain system show that intermediate horizons balance anticipation, responsiveness, and computational cost. Hardware experiments on a 6-DOF platform demonstrate consistent sim-to-real behavior without accurate inertial parameter estimation, and comparisons against dynamic optimization fabrics and model predictive control (MPC) baselines show improved success rates in dynamic clutter while preserving solve times compatible with real-time execution in the tested regimes.
Chinese Translation
机器人操纵器的实时避碰需要对意外障碍物运动做出快速反应，并提前预测以避免被近未来约束困住。全模型预测控制可以提供这种前瞻性，但其在线计算成本可能会随着预测时域长度、模型精度和活动几何约束的数量迅速增加。相反，无时域的反应方法在计算上高效，但在动态杂乱环境中可能缺乏远见。我们提出了一种任务空间递归控制器，该控制器使用短期接触一致的展开生成满足内部非穿透约束的终端运动学参考，然后仅计算向该参考平滑最小加速度过渡的第一个输入。该展开从闭环逆运动学调节法则开始，使用在膨胀的凸机器人和障碍物几何体上运行的迭代动力学求解器进行，以便机器人-障碍物接触、动态障碍物运动和自碰撞能够在不需要完全约束轨迹优化的情况下塑造终端参考。我们分析了接触不活跃的闭环，并在标准正则性假设下展示了局部指数任务空间调节。对于在展开过程中激活的接触，我们表征了相应的离散更新，并界定了移动障碍物对常规操作集的影响。对一个40自由度多链系统的仿真表明，中间时域在预见性、响应性和计算成本之间取得了平衡。在一个6自由度平台上的硬件实验展示了在没有准确惯性参数估计的情况下，模拟与现实之间的一致行为，并且与动态优化框架和模型预测控制（MPC）基线的比较显示，在动态杂乱环境中成功率有所提高，同时保持了与测试环境中的实时执行相兼容的求解时间。
cs.RO / 23 / 2607.15746
Towards Artificial Nerves: Biomimetic Optical-Fiber Tactile Sensing for Robots
迈向人工神经：仿生光纤触觉传感器在机器人中的应用
Abstract
Robotic systems increasingly demand tactile sensing that approaches the adaptability and resolution of human skin to enable dexterous manipulation and safe interaction. OptiTac is a biomimetic tactile sensor that emulates the mechanoreceptor-to-nerve architecture of human touch by pairing each mechanical pin on a soft skin with an optical fiber acting as an artificial nerve. This design demonstrates an architectural principle for routing tactile information away from the sensing surface while preserving high spatial resolution, establishing a practical route toward distributed tactile sensing in future robotic systems. By treating tactile signals as images, simple analytical methods, rather than opaque deep-learning models, are used to infer contact location, size, and shape, providing interpretable and scalable tactile intelligence. This work demonstrates how evolutionary principles from biology can guide the development of artificial nerve systems for robots, offering a pathway toward human-like tactile perception in next-generation robotic platforms. More broadly, OptiTac establishes an artificial nerve-inspired sensing framework for interpretable robotic touch and a scalable route toward future distributed tactile systems.
Chinese Translation
机器人系统日益需要接近人类皮肤适应性和分辨率的触觉传感，以实现灵巧操作和安全交互。OptiTac是一种仿生触觉传感器，通过将软皮肤上的每个机械针与作为人工神经的光纤配对，模拟人类触觉的机械感受器到神经的结构。这一设计展示了一种架构原理，可以将触觉信息从传感表面传递出去，同时保持高空间分辨率，为未来机器人系统中的分布式触觉传感建立了实际的路径。通过将触觉信号视为图像，采用简单的分析方法，而非不透明的深度学习模型，来推断接触位置、大小和形状，从而提供可解释和可扩展的触觉智能。本研究展示了生物学中的进化原理如何指导人工神经系统的开发，为下一代机器人平台提供类人触觉感知的途径。更广泛地说，OptiTac建立了一种受人工神经启发的传感框架，以实现可解释的机器人触觉，并为未来的分布式触觉系统提供可扩展的路径。
cs.RO / 24 / 2607.15758
SkillNav: Score-Level Skill Intervention for Zero-Shot Object Goal Navigation
SkillNav：零-shot目标导航的评分级技能干预
Abstract
Vision-Language Model (VLM) agents have advanced zero-shot object-goal navigation, yet single-frame reasoning leaves them without the cross-step behavioral awareness an embodied navigator requires, producing recurring failures such as dead-end stalls, in-room loops, and circuitous approaches to detected targets. Prompt-based remedies inflate token budgets across multi-submodule episodes and still struggle to encode inherently spatial signals such as angles, map cells, and viewpoint coordinates. In this paper, we propose SkillNav, an extensible behavioral skill framework for VLM-based navigation that treats the curiosity value map already maintained by modern VLM navigators as a writable substrate on which composable skills inscribe behavioral memory at zero token cost. Skills are stratified into three tiers by their level of behavioral authority, namely soft scaling for proportional reweighting, lower-bound boost for region-level guarantees, and hard override for threshold-triggered forced actions, and cooperate across tiers under a fixed composition order that establishes a predictable, declared priority among skills. This design turns capability improvement into skill registration: new behaviors plug in without retraining the VLM or disturbing existing skills, opening a path for continual refinement. A minimal prompt channel complements the score-level skills with category-level semantic hints, yielding a dual-representation design in which spatial memory lives on the map and semantic memory in short prompts. Training-free, SkillNav establishes new state-of-the-art SPL across MP3D (25.5), HM3D v0.1 (39.3), and HM3D v0.2 (43.2), improving SPL by up to 6.0 absolute over the strongest prior method, and achieves the highest Success Rate on HM3D v0.1 (69.7) and v0.2 (75.9).
Chinese Translation
视觉-语言模型（VLM）代理在零-shot目标导航方面取得了进展，但单帧推理使它们缺乏具身导航器所需的跨步骤行为意识，导致重复性失败，如死胡同停滞、房间内循环和绕道接近检测目标。基于提示的解决方案在多子模块的情境中膨胀了令牌预算，并且仍然难以编码固有的空间信号，如角度、地图单元和视点坐标。本文提出了SkillNav，一个可扩展的行为技能框架，旨在基于VLM的导航中，将现代VLM导航器已维护的好奇值图视为一个可写的基底，在其上可组合的技能以零令牌成本记录行为记忆。技能根据其行为权威级别分为三个层次，即用于比例重加权的软缩放、用于区域级保证的下限提升，以及用于阈值触发强制行为的硬覆盖，并在固定的组合顺序下跨层次协作，建立技能之间可预测的、声明的优先级。这种设计将能力提升转化为技能注册：新行为可以在不重新训练VLM或干扰现有技能的情况下接入，为持续改进开辟了道路。一个最小的提示通道与评分级技能相辅相成，提供类别级的语义提示，形成一种双重表示设计，其中空间记忆存在于地图上，语义记忆则存在于短提示中。SkillNav在无需训练的情况下，在MP3D（25.5）、HM3D v0.1（39.3）和HM3D v0.2（43.2）上建立了新的最先进的成功路径长度（SPL），相较于最强的先前方法，SPL提高了最多6.0个绝对值，并在HM3D v0.1（69.7）和v0.2（75.9）上达到了最高成功率。
cs.RO / 25 / 2607.15807
Deployment-Ready UWB Localization for Industrial Ground Robots with Automatic Anchor Calibration and Terrain-Aware Fusion
适用于工业地面机器人的部署就绪超宽带定位，具备自动锚点校准和地形感知融合
Abstract
Ultra-Wideband (UWB) ranging has become a viable option for industrial Autonomous Mobile Robot (AMR) localization due to improved accuracy and low cost. However, real-world deployments remain limited by two recurring challenges: calibrating static anchors can be time-consuming and error-prone, and integrating UWB with existing onboard sensors requires careful design to ensure robust and consistent pose estimation. Addressing these challenges, this paper presents an end-to-end pipeline that combines automatic anchor calibration with a generic multi-sensor estimator tailored to surface-bound vehicle motion. It targets existing AMR stacks in scenarios where robot pose priors are available for initialization. The calibration stage estimates anchor positions and range biases, while the localization stage fuses UWB with proprioceptive sensing in a bias-aware Extended Kalman Filter to improve consistency without extensive parameter tuning. Experiments on a commercial logistics AMR in a warehouse setting demonstrate accurate positioning indoors and across outdoor transitions, with improved consistency compared to an earlier estimator formulation. Evaluation on an independent forklift dataset further indicates transferability to other platforms. The method remains effective in test cases with limited line-of-sight and sparse anchor coverage. These results show that UWB localization can be deployed with substantially reduced manual effort while preserving the accuracy required for industrial AMRs. The collected warehouse dataset is made publicly available.
Chinese Translation
超宽带（UWB）测距因其提高的精度和低成本，已成为工业自主移动机器人（AMR）定位的可行选择。然而，现实世界的部署仍受到两个反复出现的挑战的限制：静态锚点的校准可能耗时且易出错，而将UWB与现有的车载传感器集成需要仔细设计，以确保稳健且一致的姿态估计。为了解决这些挑战，本文提出了一种端到端的流程，将自动锚点校准与针对地面车辆运动的通用多传感器估计器相结合。该方法针对在初始化时可获得机器人姿态先验的现有AMR堆栈。校准阶段估计锚点位置和范围偏差，而定位阶段则在偏差感知的扩展卡尔曼滤波器中融合UWB与自我感知传感，以提高一致性，而无需广泛的参数调整。在仓库环境中对一款商业物流AMR进行的实验表明，其在室内和户外过渡时均能实现准确定位，并且与早期的估计器公式相比，一致性得到了改善。在独立的叉车数据集上的评估进一步表明该方法可以迁移到其他平台。在有限视距和稀疏锚点覆盖的测试案例中，该方法仍然有效。这些结果表明，UWB定位可以在显著减少人工努力的同时，保持工业AMR所需的精度。收集的仓库数据集已公开提供。
cs.RO / 26 / 2607.15828
Beyond Frontiers: Scene-Anomaly Guided Autonomous Exploration
超越边界：场景异常引导的自主探索
Abstract
Autonomous exploration of unknown 3D environments is traditionally driven by coverage-maximizing geometric heuristics. However, these methods typically determine exploration targets without considering the underlying structural context. This leads to inefficient trajectories often limiting the fidelity of the final 3D reconstruction. To bridge the gap between spatial coverage and reconstruction quality, we introduce a novel paradigm: reframing exploration as a geometric anomaly minimization problem. We present SCAGE: SCene Anomaly Guided Exploration, a novel autonomous exploration framework that operates directly on unstructured 3D point clouds. Instead of blindly chasing volumetric boundaries, we equip the robot with a foundational understanding of standard indoor architecture. As the robot navigates, it continuously evaluates its live 3D observations against these learned expectations. When the incoming geometry contradicts the learned priors of a typical indoor environment, such as a fragmented wall or a partial table, the system flags these regions as scene anomalies. These geometric inconsistencies act as a guiding signal, naturally drawing the robot to investigate and resolve these structural anomalies from optimal vantage points. By actively targeting poorly reconstructed regions rather than just empty space, our approach seamlessly couples spatial discovery with high-fidelity mapping. Extensive evaluations demonstrate that SCAGE achieves superior volumetric coverage (~90% in all scenes) and higher 3D reconstruction quality compared to state-of-the-art baselines.
Chinese Translation
未知三维环境的自主探索传统上依赖于最大化覆盖的几何启发式方法。然而，这些方法通常在确定探索目标时未考虑潜在的结构上下文。这导致了效率低下的轨迹，往往限制了最终三维重建的精度。为了弥合空间覆盖与重建质量之间的差距，我们提出了一种新颖的范式：将探索重新构建为几何异常最小化问题。我们提出了SCAGE：场景异常引导探索（SCene Anomaly Guided Exploration），这是一个直接在非结构化三维点云上操作的新型自主探索框架。我们并不是盲目追逐体积边界，而是为机器人提供了对标准室内建筑的基础理解。当机器人导航时，它不断将其实时三维观察与这些学习到的期望进行评估。当输入的几何形状与典型室内环境的学习先验相矛盾时，例如破碎的墙壁或部分桌子，系统会将这些区域标记为场景异常。这些几何不一致性作为引导信号，自然吸引机器人从最佳视角去调查和解决这些结构异常。通过主动针对重建质量较差的区域，而不仅仅是空白空间，我们的方法将空间发现与高保真映射无缝结合。广泛的评估表明，SCAGE在体积覆盖率（所有场景约90%）和三维重建质量方面优于最先进的基线。
cs.RO / 27 / 2607.15863
Minimum Time Dubins Airplane Paths with Asymmetric Climb Rates
具有不对称爬升速率的最小时间Dubins飞机路径
Abstract
Dubins airplane paths approximate the limited maneuverability of fixed-wing vehicles with minimum curvature and climb rate constraints. However, the symmetric climb rate constraints result in sub-optimal paths and conservative vehicle performance. In this work, we propose asymmetric Dubins airplane paths, which consider asymmetric climb rates for climbing and descending. We revisit the time optimality conditions and show that the asymmetric flight path angle constraints preserve optimality. We show that by considering asymmetric climb rates, we can take advantage of full performance of the vehicle, reducing the minimum time by 71% for connecting randomly generated states. We also demonstrate that the added climb rate results in 2.8 times faster to find the median solution time when integrated into a sampling-based planning task on rugged terrain, due to the added feasibility. We further demonstrate the practicality of the approach with a real-world flight.
Chinese Translation
Dubins飞机路径在最小曲率和爬升速率约束下，近似固定翼飞行器的有限机动性。然而，对称的爬升速率约束导致了次优路径和保守的飞行器性能。在本研究中，我们提出了不对称Dubins飞机路径，考虑了爬升和下降过程中的不对称爬升速率。我们重新审视了时间最优性条件，并表明不对称飞行路径角约束保持了最优性。我们展示了通过考虑不对称爬升速率，可以充分利用飞行器的性能，将连接随机生成状态的最小时间减少71%。我们还证明，在崎岖地形的基于采样的规划任务中，增加的爬升速率使得找到中位解的时间提高了2.8倍，因其增加了可行性。我们进一步通过实际飞行演示了该方法的实用性。
cs.RO / 28 / 2607.15880
Dynamics-Aware Meta-Imitation for Generalization to Unseen Robotic Manipulation
动态感知元模仿以实现对未见机器人操作的泛化
Abstract
Imitation Learning aims to learn skills from extensive observations and demonstrations for robots, so it suffers from data scarcity and environment generalization. The existing methods predominantly focus on imitation from in-domain tasks and consequently struggle with generalization to unseen tasks. To bridge this generalization gap, we propose the \textbf{D}ynamics-\textbf{A}ware \textbf{M}eta-\textbf{I}mitation (DAMI) framework. By integrating meta-learning to construct a shared skill space, DAMI equips agents for rapid adaptation to novel tasks. We introduce the Visual-Motor Trajectory (VMT) module to capture complex spatio-temporal dynamics within the task latent space. Furthermore, we propose the Unpaired Unified Task (U2T) block to fuse unstructured multimodal observations. To coordinate these representations, we integrate a Task-Conditioned Feature Modulation (TCFM) mechanism customized for modulating low-level 3D features. By capturing intrinsic dynamics from a random complete reference demonstration, our framework learns the underlying task logic rather than memorizing static cues, ensuring effective generalization. Extensive experiments in both simulation and real-world settings demonstrate that our approach outperforms state-of-the-art baselines regarding direct inference on seen tasks and adaptation to unseen tasks via few-shot fine-tuning.
Chinese Translation
模仿学习旨在通过大量观察和示范来学习机器人技能，因此面临数据稀缺和环境泛化的问题。现有方法主要集中在从领域内任务进行模仿，因此在对未见任务的泛化方面存在困难。为了解决这一泛化差距，我们提出了 extbf{D}ynamics- extbf{A}ware extbf{M}eta- extbf{I}mitation (DAMI)框架。通过整合元学习构建共享技能空间，DAMI使代理能够快速适应新任务。我们引入了视觉-运动轨迹（Visual-Motor Trajectory, VMT）模块，以捕捉任务潜在空间中的复杂时空动态。此外，我们提出了无配对统一任务（Unpaired Unified Task, U2T）模块，以融合非结构化的多模态观察。为了协调这些表示，我们集成了任务条件特征调制（Task-Conditioned Feature Modulation, TCFM）机制，专门用于调制低级3D特征。通过从随机完整参考示范中捕捉内在动态，我们的框架学习潜在任务逻辑，而不是记忆静态线索，从而确保有效的泛化。在仿真和现实环境中的大量实验表明，我们的方法在已见任务的直接推理和通过少量样本微调适应未见任务方面优于最先进的基线。
cs.RO / 29 / 2607.15935
Learning Reach-Avoid Task with Reinforcement Learning: Vectorized Simulation and Benchmark
使用强化学习学习达避任务：向量化仿真与基准测试
Abstract
Deep reinforcement learning (DRL) has a longstanding tradition in addressing the reach-avoid task problem, especially for controlling robotic arms. While this task serves as a baseline environment within the research community, the ability of DRL to effectively learn the each-avoid task in complex and realistic scenarios beyond simplified and restricted tabletop settings remains uncertain. In this paper, we present, for the first time, a comprehensive benchmark for the reachavoid task that accurately captures real-world complexities without simplifications. We demonstrate a diverse range of settings for robotic arm reach-avoid task, which can be used for evaluating DRL research. We achieved this by utilizing the MuJoCo MJX physics engine and parallelizing both the simulation environment and DRL algorithms using the Brax library. We achieved state-of-the-art results with success rates of 96.1% (UR5e) and 98.8% (Franka Emika Robot) for the reach task and 86.8% (UR5e) and 95.2% (Franka) for the static reachavoid task. Our results indicate that while in previous works DRL agents could solve, for example, a reach task in a simplified setting perfectly, their agents performance collapses when evaluated in realistic scenarios. Overall, this work identifies that additional research is still required to claim the successful resolution of the robotic arm reach-avoid task using DRL. The environment and benchmarking code is available as open source at the following link
Chinese Translation
深度强化学习（DRL）在解决达避任务问题方面有着悠久的传统，特别是在控制机器人手臂方面。尽管该任务在研究界作为基准环境，但DRL在复杂和真实场景中有效学习达避任务的能力仍然不确定，这些场景超出了简化和受限的桌面设置。在本文中，我们首次提出了一个全面的达避任务基准，准确捕捉现实世界的复杂性而不进行简化。我们展示了一系列多样化的机器人手臂达避任务设置，可用于评估DRL研究。我们通过利用MuJoCo MJX物理引擎，并使用Brax库对仿真环境和DRL算法进行并行化，达成了这一目标。我们在达任务上取得了96.1%（UR5e）和98.8%（Franka Emika Robot）的成功率，在静态达避任务上取得了86.8%（UR5e）和95.2%（Franka）的成功率。我们的结果表明，尽管在之前的工作中，DRL代理能够在简化设置中完美解决达任务，但在真实场景中评估时其性能却崩溃。总体而言，这项工作表明，仍需进行更多研究以声称成功解决机器人手臂达避任务的DRL方法。环境和基准测试代码可在以下链接以开源形式获取。
cs.RO / 30 / 2607.15974
Embodied Active Learning under Limited Annotation and Navigation Budget for Object Detection
在有限标注和导航预算下的具身主动学习用于目标检测
Abstract
This paper studies how to adapt a computer vision object detector to an unknown environment under both a robot navigation time and annotation budget constraint. Our approach selects informative robot trajectories and image samples to retrain the detector, explicitly targeting its failure cases. Formally, the approach is an embodied variant of batch active learning, where at each round an agent has a limited navigation budget to collect candidate samples and a limited annotation budget for the most relevant images. We leverage spatial consistency to identify images with inconsistent labels, which are likely to provide the greatest improvement to the vision model. We evaluate the approach using different active learning objectives on large scenes from the AI2-THOR simulator and on a real-world setup using a Boston Dynamics Spot robot with the real-time object detector YOLOv5. Through comparison against several baselines, our experimental results show that spatial inconsistency helps guide the agent and select relevant images without external supervision, achieving the highest detection accuracy at the end of the adaptation process under the same budget. The open-source project can be found at https://mkabouri.github.io/embodied-active-learning-od
Chinese Translation
本文研究如何在机器人导航时间和标注预算限制下，将计算机视觉目标检测器适应于未知环境。我们的方法选择信息丰富的机器人轨迹和图像样本来重新训练检测器，明确针对其失败案例。形式上，该方法是批量主动学习的具身变体，其中在每一轮中，代理有一个有限的导航预算来收集候选样本，并且对于最相关的图像有一个有限的标注预算。我们利用空间一致性来识别标签不一致的图像，这些图像可能为视觉模型提供最大的改进。我们在AI2-THOR模拟器的大场景和使用波士顿动力Spot机器人及实时目标检测器YOLOv5的真实场景中，使用不同的主动学习目标对该方法进行了评估。通过与多个基线的比较，我们的实验结果表明，空间不一致性有助于引导代理并选择相关图像，而无需外部监督，在相同预算下实现了适应过程结束时的最高检测准确率。开源项目可在https://mkabouri.github.io/embodied-active-learning-od找到。
cs.RO / 31 / 2607.15982
Data and Learning Where it Matters for Contact-Rich Manipulation
在接触丰富的操作中，数据和学习的重要性
Abstract
Learned policies trained end-to-end on large datasets often remain brittle in high-precision tasks and struggle with generalization. We find that these limitations largely stem from a lack of structure and focus in data collection. Our key insight is to leverage dense data collection only for the critical segment of contact-rich tasks and to rely on traditional planning during simple free-space motion. We propose an automated data-collection scheme in combination with offline deep reinforcement learning for the critical segment of the task, eliminating reliance on a teleoperator's skill and on online policy updates. Across four challenging real-world tasks, using only 2 to 2.5 hours of autonomous data collection, we achieve an average success rate of 96%, compared to the strongest baseline at 55%. Notably, performance remains high in out-of-distribution scenarios where end-to-end approaches struggle. Our results pave the way for targeted data collection for contact-rich tasks and for high success rates in precision applications.
Chinese Translation
在大型数据集上端到端训练的学习策略在高精度任务中往往表现脆弱，并且在泛化能力上存在困难。我们发现，这些局限性主要源于数据收集缺乏结构和重点。我们的关键见解是，仅在接触丰富任务的关键环节利用密集数据收集，而在简单的自由空间运动中依赖传统规划。我们提出了一种自动化数据收集方案，结合离线深度强化学习，专注于任务的关键环节，消除了对遥控操作员技能和在线策略更新的依赖。在四个具有挑战性的真实世界任务中，仅使用2到2.5小时的自主数据收集，我们实现了96%的平均成功率，而最强基线仅为55%。值得注意的是，在端到端方法表现不佳的分布外场景中，性能仍然保持较高水平。我们的结果为接触丰富任务的有针对性数据收集以及在精密应用中的高成功率铺平了道路。
cs.RO / 32 / 2607.16002
A Morphing-Designed Hexarotor Prototype combining Practical Resilience and Efficiency
一种结合实用韧性和效率的变形设计六旋翼原型
Abstract
This work demonstrates experimentally the existence of a hexarotor prototype, termed Opti-Hexa, that simultaneously achieves practical resilience to single-propeller failures and energy efficiency comparable to a standard Star-shaped prototype with the same size, weight, hardware and software. Leveraging a novel open-source morphing platform, we investigate the trade-offs across a continuous range of geometries by varying the angles between adjacent propellers. We study practical efficiency through a data-fitted empirical power model and evaluate practical resilience by comparing the position accuracy and rotational kinetic energy during failure to those observed under nominal hovering conditions. Our experiments confirm the existence of a geometric viability region for this specific morphing platform, where resilience is ensured without the aerodynamic efficiency losses typically associated with practically resilient designs found in the state of the art. The complete hardware and software of the morphing platform are released to support further research.
Chinese Translation
本研究通过实验展示了一种名为Opti-Hexa的六旋翼原型，其能够同时实现对单个螺旋桨故障的实用韧性和与同尺寸、重量、硬件及软件的标准星形原型相当的能源效率。利用一种新颖的开源变形平台，我们通过改变相邻螺旋桨之间的角度，研究了在连续几何形状范围内的权衡。我们通过数据拟合的经验功率模型研究了实用效率，并通过比较故障期间的位置精度和旋转动能与名义悬停条件下的观察结果来评估实用韧性。我们的实验确认了该特定变形平台的几何可行性区域的存在，在该区域内，韧性得以保证，而不会出现与当前最先进的实用韧性设计通常相关的气动效率损失。完整的变形平台硬件和软件已发布，以支持进一步的研究。
cs.RO / 33 / 2607.16095
Let the Body Follow: Coupled Egocentric Control for Whole-Body Robot Teleoperation
让身体跟随：全身机器人遥操作的耦合自我中心控制
Abstract
Whole-body teleoperation requires users to coordinate perception, manipulation, posture, and mobility across multiple robot components. This coordination is difficult because users must simultaneously control the robot's head, arms, torso, and base while maintaining task awareness and avoiding kinematic or environmental constraints. In this paper, we propose coupled egocentric control, a body-following teleoperation approach in which the robot's torso and base automatically respond to the operator's head and arm motions. Rather than requiring explicit touchpad commands for every torso or base adjustment, the system lets users focus on gaze and hand control: head pitch adjusts torso height, head yaw drives base rotation, end-effector height adjusts torso motion, and end-effector workspace boundaries trigger base translation. We evaluate this approach in a user study on whole-body teleoperation of a TIAGo mobile manipulator for home-care-inspired tasks. Compared with a baseline hybrid interface, coupled egocentric control improves object manipulation efficiency, reduces button-based control effort and arm singularities, lowers mental demand and overall workload, and increases ease of use, ease of learning, confidence, and user preference for torso and base control.
Chinese Translation
全身遥操作要求用户协调多个机器人组件的感知、操作、姿态和移动。这种协调很困难，因为用户必须同时控制机器人的头部、手臂、躯干和底座，同时保持任务意识并避免运动学或环境约束。在本文中，我们提出了耦合自我中心控制，这是一种身体跟随的遥操作方法，其中机器人的躯干和底座自动响应操作员的头部和手臂动作。该系统不再要求用户为每次躯干或底座的调整提供明确的触控板命令，而是让用户专注于视线和手部控制：头部俯仰调整躯干高度，头部偏航驱动底座旋转，末端执行器高度调整躯干运动，末端执行器工作空间边界触发底座平移。我们在一项用户研究中评估了这种方法，该研究针对TIAGo移动操控器在家庭护理启发任务中的全身遥操作。与基线混合接口相比，耦合自我中心控制提高了物体操作效率，减少了基于按钮的控制努力和手臂奇异性，降低了心理需求和整体工作负荷，并增加了使用的便利性、学习的容易性、自信心和用户对躯干和底座控制的偏好。
cs.RO / 34 / 2607.16123
BayesContact: Uncertain Pose Estimation via Visuo-Tactile Proposals and Simulation-based Inference
BayesContact：通过视觉-触觉提议和基于仿真的推断进行不确定姿态估计
Abstract
Contact-rich manipulation requires pose estimates that are often more accurate than what depth-only sensing provides. Existing methods, relying on vision and contact, employ costly offline training procedures that need to be retrained for new environments and geometries. We propose BayesContact, a Simulation-Based Inference framework for visuo-tactile pose estimation in peg-in-hole insertion. BayesContact maintains a particle belief over object pose and fuses depth observations with force/torque-derived contact evidence. We employ simulation based forward models to approximate these observation likelihoods. For each pose hypothesis, a renderer predicts depth measurements and a physics simulator predicts contact outcomes under guarded probing actions; both are scored against real observations to update the belief. The resulting multimodal belief also enables information-gain-based probing for active disambiguation. Across simulated geometries and real-robot experiments, BayesContact improves pose observability and insertion success over vision-only inference by 30%
Chinese Translation
接触丰富的操作需要比仅依赖深度传感器提供的姿态估计更为准确的结果。现有方法依赖于视觉和接触，采用昂贵的离线训练程序，这些程序需要在新的环境和几何形状下重新训练。我们提出了BayesContact，一个用于插销-孔插入的视觉-触觉姿态估计的基于仿真的推断框架。BayesContact在物体姿态上维持粒子信念，并将深度观测与基于力/扭矩的接触证据融合。我们采用基于仿真的前向模型来近似这些观测的似然性。对于每个姿态假设，渲染器预测深度测量，物理模拟器在受控探测动作下预测接触结果；这两者都与真实观测进行评分，以更新信念。最终生成的多模态信念还支持基于信息增益的探测，以实现主动消歧。在模拟几何形状和真实机器人实验中，BayesContact在姿态可观测性和插入成功率上比仅依赖视觉的推断提高了30%。
cs.RO / 35 / 2607.16143
A New Implementation of NeoSLAM and a Comparative Evaluation with RatSLAM
NeoSLAM的新实现及其与RatSLAM的比较评估
Abstract
This paper presents a new implementation of the NeoSLAM algorithm. The proposed version is a complete rewrite of NeoSLAM into a modular architecture using modern frameworks that, together, enable real-time execution with minimal discarding of input data. This work also provides a comparative evaluation between NeoSLAM and RatSLAM across three datasets under varying environmental conditions. The experimental results highlight differences in mapping consistency and trajectory reconstruction, demonstrating the effectiveness and practical applicability of the proposed ROS2-based implementation. The results indicate that the new NeoSLAM outperforms the original in terms of processing throughput for real-time applications and achieves comparable performance to RatSLAM in terms of map reconstruction across the evaluated datasets.
Chinese Translation
本文提出了一种NeoSLAM算法的新实现。所提版本是对NeoSLAM的完整重写，采用模块化架构和现代框架，使得能够在实时执行中最小化输入数据的丢弃。本文还提供了在不同环境条件下，NeoSLAM与RatSLAM在三个数据集上的比较评估。实验结果突出了映射一致性和轨迹重建的差异，展示了所提基于ROS2的实现的有效性和实际应用性。结果表明，新版本的NeoSLAM在实时应用中的处理吞吐量上优于原始版本，并且在评估的数据集中，其地图重建性能与RatSLAM相当。
cs.RO / 36 / 2607.16146
VTLoc: Learning-based Tactile Contact Localization in Visual Point Clouds
VTLoc：基于学习的视觉点云触觉接触定位
Abstract
Vision and touch are complementary modalities essential for robotic perception and manipulation. While vision provides global object context, touch offers precise local information at contact points. Integrating these modalities for contact localization, i.e., predicting the location of touch on an object's surface, poses significant challenges due to the need for accurate spatial alignment between tactile data and visual geometry. To address this challenge, we propose VTLoc, a novel visual-tactile framework that localizes contact points from tactile readings using a 3D point cloud as visual input. VTLoc introduces two key components: a geometric multi-modal alignment module, which reconstructs a pseudo-point cloud from fused visual-tactile features and aligns it with the visual point cloud to enforce spatial consistencies across modalities; and an iterative localizing updater, which iteratively refines the predicted contact location using fused visual-tactile features. Evaluated on a new benchmark of 100 real-world objects, VTLoc improves single-touch contact localization by reducing local-to-global correspondence ambiguity.
Chinese Translation
视觉和触觉是机器人感知和操作中不可或缺的互补模态。视觉提供了全局物体上下文，而触觉则在接触点提供了精确的局部信息。将这两种模态整合以进行接触定位，即预测触摸在物体表面的位置，面临着显著的挑战，因为需要在触觉数据和视觉几何之间实现准确的空间对齐。为了解决这一挑战，我们提出了VTLoc，一个新颖的视觉-触觉框架，它利用3D点云作为视觉输入，从触觉读数中定位接触点。VTLoc引入了两个关键组件：几何多模态对齐模块，该模块从融合的视觉-触觉特征重建伪点云，并将其与视觉点云对齐，以在模态之间强制执行空间一致性；以及迭代定位更新器，该更新器使用融合的视觉-触觉特征迭代精炼预测的接触位置。在对100个真实物体的新基准进行评估时，VTLoc通过减少局部到全局对应关系的模糊性，提高了单触摸接触定位的准确性。
cs.RO / 37 / 2607.16173
Vision-Language-Motion Maps: An Open-Vocabulary, Uncertainty-Aware, Queryable Motion Attribute for 3D Scene Maps
视觉-语言-运动地图：一种开放词汇、考虑不确定性的可查询运动属性用于 3D 场景地图
Abstract
Open-vocabulary 3D maps let robots answer language queries about what and where, but they assume a static world and cannot answer queries about how scene elements behave. We introduce Vision-Language-Motion Maps (VLMM), an open-vocabulary, natural-language-queryable 3D map in which each element carries a fused motion attribute: a VLM/LLM semantic movability prior combined with geometrically observed cross-frame motion, together with a per-element uncertainty. Queries reduce to attribute filters that distinguish what has been seen to move, what could move but has not, and what stays still. On a controlled simulator benchmark with exact ground truth (AI2-THOR, three scene types) we show through ablation that the schema fields are non-substitutable: a semantic-only baseline fails motion queries even with strong features, and neither motion field substitutes for the other (the prior cannot answer "what is moving," observed motion cannot answer "what could move"). On real dynamic RGB-D (TUM and Bonn, six sequences) we show the uncertainty channel-our key difference from prior fused-motion work-consistently improves moving-vs-static average precision and reduces false motion flags, and is robust to estimated (noisy) poses. The raw confidence is not calibrated, but post-hoc isotonic calibration reaches an expected calibration error of 0.10. VLMM is a representation contribution: the closest prior maps each lack at least one of the four properties-open-vocabulary, language-queryable, fused prior-and-observed motion, and per-element uncertainty-that our combination provides.
Chinese Translation
开放词汇的 3D 地图使机器人能够回答关于“什么”和“哪里”的语言查询，但它们假设世界是静态的，无法回答关于场景元素如何行为的查询。我们引入了视觉-语言-运动地图（Vision-Language-Motion Maps, VLMM），这是一种开放词汇、可用自然语言查询的 3D 地图，其中每个元素都携带一个融合的运动属性：结合了 VLM/LLM 语义可移动性先验和几何观察到的跨帧运动，以及每个元素的不确定性。查询简化为属性过滤器，以区分已观察到的运动、可能运动但尚未运动的元素，以及静止的元素。在一个具有精确真实值的受控模拟基准（AI2-THOR，三种场景类型）中，我们通过消融实验展示了模式字段是不可替代的：仅语义的基线即使具有强特征也无法满足运动查询，而两个运动字段也无法相互替代（先验无法回答“什么在移动”，观察到的运动无法回答“什么可能移动”）。在真实动态 RGB-D 数据集（TUM 和 Bonn，六个序列）上，我们展示了不确定性通道——这是我们与先前融合运动工作的关键区别——持续提高了移动与静态的平均精度，并减少了错误的运动标志，并且对估计的（噪声）姿态具有鲁棒性。原始置信度未经过校准，但事后等距校准达到了 0.10 的预期校准误差。VLMM 是一种表示贡献：最接近的先前地图每个都缺乏我们组合所提供的四个属性中的至少一个——开放词汇、可语言查询、融合的先验和观察到的运动，以及每个元素的不确定性。
cs.RO / 38 / 2607.16187
Handroid: Bridging Dexterous Hand and Humanoid
Handroid：连接灵巧手与类人机器人
Abstract
Dexterous hands and humanoid robots are typically developed as distinct embodiments: the former enable contact-rich manipulation at the object scale, whereas the latter provide mobility and whole-body interaction in human-centered environments. We introduce \textbf{Handroid}, a desktop-scale dual-embodiment robot that integrates both capabilities within a single reconfigurable platform. Handroid reuses one 27-DoF electromechanical body as either a dexterous hand or a desktop humanoid, measuring 0.33 m in height and 2.05 kg in weight. In the dexterous hand embodiment, 20 DoFs form an anthropomorphic hand closely matching the kinematic structure of the human hand. In the humanoid embodiment, the same articulated modules are reconfigured into a humanoid with a head, arms, and legs, including a 12-DoF lower-limb structure for locomotion and whole-body motion. Handroid further provides a unified control and learning framework supporting hand teleoperation, dexterous grasping, in-hand manipulation, humanoid locomotion, gait generation, and interactive motion authoring. We validate the platform through real-world dexterous manipulation, reinforcement-learning-based locomotion, keyframe motion deployment, and a long-horizon task involving embodiment reconfiguration, locomotion, docking, and dexterous pick-and-place. These results position Handroid as a compact and reproducible platform for advancing morphology-reconfigurable robotics and cross-embodiment robot learning.
Chinese Translation
灵巧手和类人机器人通常被开发为不同的体现形式：前者能够在物体尺度上进行丰富的接触操控，而后者则在以人为中心的环境中提供移动性和全身交互。我们介绍了 extbf{Handroid}，一种桌面规模的双体现机器人，它在一个可重构的平台上集成了这两种能力。Handroid重用一个27自由度（DoF）的机电身体，既可以作为灵巧手，也可以作为桌面类人机器人，身高为0.33米，重量为2.05千克。在灵巧手的体现中，20个自由度形成一个与人手的运动学结构紧密匹配的人形手。在类人机器人的体现中，相同的关节模块被重新配置为一个具有头部、手臂和腿部的类人机器人，包括一个用于移动和全身运动的12自由度下肢结构。Handroid进一步提供了一个统一的控制和学习框架，支持手部遥操作、灵巧抓取、手内操控、类人机器人移动、步态生成和交互运动创作。我们通过现实世界中的灵巧操控、基于强化学习的移动、关键帧运动部署，以及涉及体现重构、移动、对接和灵巧拾取与放置的长时间任务来验证该平台。这些结果使Handroid成为一个紧凑且可重复的平台，有助于推进形态可重构机器人技术和跨体现机器人学习。
cs.CV / 1 / 2607.15288
An Empirical Study of Handcrafted Feature Learning and Convolutional Neural Networks for Facial Expression Recognition
手工特征学习与卷积神经网络在面部表情识别中的实证研究
Abstract
Facial expression recognition is an important computer vision task with applications in human--computer interaction, mental health monitoring, driver alert systems, and behavioral analysis. While convolutional neural networks (CNNs) dominate modern facial expression recognition, handcrafted feature descriptors such as Histogram of Oriented Gradients (HOG) and Local Binary Patterns (LBP) remain useful classical baselines. This study compares HOG with Support Vector Machine (SVM), LBP with Logistic Regression, and a lightweight CNN across three facial expression datasets: FER-2013, CK+, and KDEF. The results show that CNNs achieve the best overall performance, particularly on more complex data, while HOG performs strongly in controlled environments. LBP performs poorly across all datasets. The study highlights that dataset complexity significantly affects performance and that robust feature learning is essential for real-world facial expression recognition.
Chinese Translation
面部表情识别是一个重要的计算机视觉任务，广泛应用于人机交互、心理健康监测、驾驶警示系统和行为分析等领域。尽管卷积神经网络（CNN）在现代面部表情识别中占据主导地位，但手工特征描述符如方向梯度直方图（HOG）和局部二值模式（LBP）仍然是有用的经典基线。本研究比较了HOG与支持向量机（SVM）、LBP与逻辑回归，以及一种轻量级CNN在三个面部表情数据集上的表现：FER-2013、CK+和KDEF。结果表明，CNN在整体性能上表现最佳，尤其是在更复杂的数据上，而HOG在受控环境中表现强劲。LBP在所有数据集上的表现均较差。研究强调数据集的复杂性显著影响性能，并指出稳健的特征学习对于现实世界的面部表情识别至关重要。
cs.CV / 2 / 2607.15321
Rethinking the Readout: Unlocking Video Backbones for AI-Generated Video Detection
重新思考读出：解锁视频骨干网络以进行AI生成视频检测
Abstract
AI-generated videos (AIGVs) typically contain subtle temporal artifacts that arise from inter-frame inconsistencies rather than within individual frames. A detector that captures such artifacts should therefore benefit from video pretrained backbones over image only ones. In practice, however, video backbones with standard global readouts often fail to outperform strong image pretrained probes on AIGV benchmarks. We attribute this gap to excessive spatiotemporal aggregation in the readout. Video pretrained backbones tend to compress each frame into a single global descriptor. This compression suppresses local patch level temporal dynamics and discards inter patch relations, which are precisely the cues that AIGV detection most reliably depends on. Based on this, we propose Velocity Gated Patch Velocity Profiling (V-PVP), a lightweight readout that replaces only the aggregation layer with two parallel streams over the patch velocity field, adding only about $0.5$M trainable parameters. V-PVP serves as a general plug-and-play module that consistently improves performance across diverse video backbones under both end-to-end fine-tuning and linear probing settings. Our method reaches \textbf{95.28} AUC on AIGVDBench while keeping the backbone fully frozen. The results show that simply replacing the aggregation layer reactivates the temporal potential of frozen video backbones, restoring their advantage on AIGV detection. Code is available at https://anonymous.4open.science/r/PVP-81B3/.
Chinese Translation
AI生成的视频（AIGVs）通常包含微妙的时间伪影，这些伪影源于帧间不一致性，而非单个帧内的因素。因此，能够捕捉这些伪影的检测器应当受益于视频预训练的骨干网络，而非仅图像的骨干网络。然而，在实践中，使用标准全局读出的video骨干网络在AIGV基准测试中往往未能超越强大的图像预训练探测器。我们将这一差距归因于读出过程中过度的时空聚合。视频预训练的骨干网络倾向于将每一帧压缩为一个单一的全局描述符。这种压缩抑制了局部补丁级别的时间动态，并丢弃了补丁间的关系，而这些恰恰是AIGV检测最可靠依赖的线索。基于此，我们提出了速度门控补丁速度剖面（Velocity Gated Patch Velocity Profiling, V-PVP），这是一种轻量级的读出方法，仅用两个并行流替代聚合层，作用于补丁速度场，增加的可训练参数约为0.5M。V-PVP作为一个通用的即插即用模块，在端到端微调和线性探测设置下，始终能够提升多种视频骨干网络的性能。我们的方法在AIGVDBench上达到了95.28的AUC，同时保持骨干网络完全冻结。结果表明，简单地替换聚合层重新激活了冻结视频骨干网络的时间潜力，恢复了其在AIGV检测中的优势。代码可在 https://anonymous.4open.science/r/PVP-81B3/ 获取。
cs.CV / 3 / 2607.15331
Training-Free Open-Vocabulary 3D Point-Cloud Segmentation on the Generalized Few-Shot Benchmark
无训练开放词汇的三维点云分割在广义少样本基准上的应用
Abstract
Generalized few-shot 3D point-cloud segmentation (GFS-PCS) asks a model to segment a scene into many base classes seen at training time and a set of novel classes. The state of the art reaches novel classes by reconciling a dense but noisy 3D vision-language prior with the few-shot support, but it pays for this with base 3D labels, per-episode training, and the support annotations themselves. We ask how far the same reconciliation can go with none of these: no training, no 3D labels, and not even the few-shot support. We pair a frozen 3D vision-language model (RegionPLC) as a dense prior with a frozen promptable concept segmenter (SAM3), prompted by the bare novel class names and lifted from posed RGB views, and reconcile the two by cross-view consistency: a point becomes novel only when enough of the views that see it agree. On the ScanNet200 GFS-PCS benchmark this fully training-free, open-vocabulary pipeline improves novel mIoU by +2.6 over the training-free dense prior while holding base accuracy within 0.5, and recovers a third (33%) of the novel-class gap to the trained state of the art that uses far more supervision. We further show that injecting the few-shot support into the pipeline, as a fusion gate and as a prototypical dense classifier, adds nothing over consistency alone and in fact degrades it through the classifier, which is why the method needs no support at all. On the harder ScanNet++ benchmark, where the dense prior is far weaker on novel classes, the same pipeline nearly doubles novel mIoU (+15.7, from 16.2 to 31.9) at a 1.7 base cost, lifting the harmonic mean from 21.5 to 31.1
Chinese Translation
广义少样本三维点云分割（GFS-PCS）要求模型将场景分割为训练时见过的多个基础类别和一组新类别。当前的最先进技术通过将密集但噪声较大的三维视觉-语言先验与少样本支持进行调和，来实现新类别的分割，但这需要基础三维标签、每个回合的训练以及支持注释本身。我们探讨在没有这些条件下，调和的效果能达到什么程度：没有训练、没有三维标签，甚至没有少样本支持。我们将一个冻结的三维视觉-语言模型（RegionPLC）作为密集先验，与一个冻结的可提示概念分割器（SAM3）配对，后者通过裸露的新类别名称和从指定的RGB视图中提取的方式进行提示，并通过跨视图一致性进行调和：一个点只有在足够多的视图看到它并达成一致时，才被视为新类别。在ScanNet200 GFS-PCS基准上，这一完全无训练的开放词汇管道在保持基础准确率在0.5以内的同时，将新类别的mIoU提高了+2.6，相较于无训练的密集先验，并且恢复了与使用更多监督的训练最先进技术之间三分之一（33%）的新类别差距。我们进一步表明，将少样本支持注入管道，作为融合门和原型密集分类器，除了通过一致性本身外并没有增加任何效果，实际上通过分类器降低了效果，这就是该方法完全不需要支持的原因。在更具挑战性的ScanNet++基准上，密集先验在新类别上的表现远不如前述方法，该管道几乎将新类别的mIoU翻倍（+15.7，从16.2提升至31.9），基础成本为1.7，使调和平均值从21.5提升至31.1。
cs.CV / 4 / 2607.15374
Reasoning-Guided Part-Level Visual Grounding via Reinforcement Learning
基于强化学习的推理引导部件级视觉定位
Abstract
Multimodal large language models (MLLMs) ground whole objects well from free-form language queries, but they struggle when the query names a part rather than the object. We trace this to a missing object-part hierarchy, since parts are localized in the same single step used for objects. We propose Object-Part Hierarchical Reflective Grounding (OP-HRG), a coarse-to-fine reasoning-guided grounding strategy that first localizes the parent object and then the part within it. A self-check then reflects on the result, with an extension to re-encode the predicted crop to inspect the region it is correcting. We introduce a part-aware GRPO framework to train our pipeline with stage-wise rewards. A 4B model trained this way outperforms 7B grounding LLMs and SAM3 across PascalPart, PartImageNet, and InstructPart, and transfers to reasoning segmentation.
Chinese Translation
多模态大型语言模型（MLLMs）能够从自由形式的语言查询中很好地定位整个物体，但当查询指定部件而非物体时，它们却表现不佳。我们将这一问题归因于缺失的物体-部件层级结构，因为部件是在与物体相同的单一步骤中进行定位的。我们提出了物体-部件层级反射定位（Object-Part Hierarchical Reflective Grounding, OP-HRG），这是一种粗到细的推理引导定位策略，首先定位父物体，然后在其中定位部件。接下来进行自我检查，以反思结果，并扩展到重新编码预测的裁剪区域，以检查其修正的区域。我们引入了一种部件感知的GRPO框架，以阶段性奖励训练我们的管道。通过这种方式训练的4B模型在PascalPart、PartImageNet和InstructPart上超越了7B的定位LLMs和SAM3，并能够迁移到推理分割任务。
cs.CV / 5 / 2607.15396
Partial Information Decomposition as a Multi-Contrast 3D MRI Selection Strategy for Resource-Constrained Deep Neural Network Training in Brain Tumor Segmentation
作为资源受限深度神经网络训练的多对比3D MRI选择策略的部分信息分解在脑肿瘤分割中的应用
Abstract
Multi-contrast 3D MRI segmentation can be computationally demanding when all available sequences are used. We evaluate a pre-training Partial Information Decomposition framework that ranks input pairs according to their redundant, unique, and synergistic information about regional tumor burden and selects the highest-ranked pair for downstream training. Applied to T1n, T1c, T2w, and T2-FLAIR MRI, the framework selected T1c+T2-FLAIR. We then trained eleven architecturally identical lightweight 3D U-Nets using different input configurations. On an independent test cohort, T1c+T2-FLAIR was the strongest two-input configuration and ranked second overall in mean Dice (0.676 versus 0.687 for all four inputs). Independent Shapley analysis on the full-input model also identified T2-FLAIR and T1c as the most influential inputs and their pairwise interaction as the strongest. These findings demonstrate the practical value of PID based pre-training selection for identifying compact, informative MRI input sets before costly 3D model development.
Chinese Translation
在使用所有可用序列时，多对比3D MRI分割可能会计算上非常耗费资源。我们评估了一种预训练的部分信息分解框架，该框架根据输入对关于区域肿瘤负担的冗余、独特和协同信息对其进行排名，并选择排名最高的对进行下游训练。该框架应用于T1n、T1c、T2w和T2-FLAIR MRI，选择了T1c+T2-FLAIR。随后，我们使用不同的输入配置训练了十一种架构相同的轻量级3D U-Net。在一个独立的测试队列中，T1c+T2-FLAIR是最强的双输入配置，并在平均Dice得分中排名第二（0.676对比于所有四个输入的0.687）。对全输入模型的独立Shapley分析也确认了T2-FLAIR和T1c是最具影响力的输入，其成对交互作用最强。这些发现展示了基于PID的预训练选择在识别紧凑且信息丰富的MRI输入集方面的实际价值，从而为成本高昂的3D模型开发提供支持。
cs.CV / 6 / 2607.15400
Unsupervised Keypoints for Real-Time Fall Detection: Comparative Analysis Under Real-world Conditions with Predictive Bandwidth Reduction
无监督关键点用于实时跌倒检测：在真实世界条件下的比较分析与预测带宽降低
Abstract
Falls among older adults are a major safety challenge, but continuous monitoring is difficult to sustain. Video captures fall-related posture and motion, yet deployment is limited by privacy, computation, and bandwidth. Supervised pose estimation is anatomically interpretable but vulnerable to occlusion and partial body visibility. We propose a privacy-preserving framework that replaces RGB transmission with compact motion representations based on unsupervised keypoints and predictive temporal modeling. Local processing performs segmentation and keypoint extraction; variational recurrent prediction and sequence classification then detect falls from observed and forecasted motion. We evaluate the framework on the UR Fall Detection and Human Fall datasets using random, subject-disjoint, and occlusion-based splits. Under random splits, neither representation consistently dominates, suggesting that standard protocols may hide meaningful differences. Under subject-disjoint evaluation, supervised keypoints show a statistically significant advantage, but performance varies by subject: they perform better when anatomical landmarks are visible, whereas unsupervised keypoints are more robust to occlusion and partial visibility, though they produce more false positives for complex activities. Under occlusion-based evaluation, supervised keypoints miss nearly half of all falls, while unsupervised keypoints retain strong sensitivity and substantially outperform them. Their anatomical independence allows spatial anchors to adapt to visible body structure rather than fail on absent landmarks. The gap widens under bandwidth constraints, where supervised localization errors compound through the temporal model. These findings show that representation choice should reflect expected visual conditions and that unsupervised keypoints offer an advantage when body visibility is compromised.
Chinese Translation
老年人跌倒是一个主要的安全挑战，但持续监测难以维持。视频捕捉与跌倒相关的姿势和运动，但由于隐私、计算和带宽的限制，部署受到限制。监督姿态估计在解剖上可解释，但容易受到遮挡和部分身体可见性的影响。我们提出了一种保护隐私的框架，通过基于无监督关键点和预测时间建模的紧凑运动表示替代RGB传输。局部处理执行分割和关键点提取；变分递归预测和序列分类随后从观察到的和预测的运动中检测跌倒。我们在UR跌倒检测和人类跌倒数据集上评估该框架，使用随机、主体不重叠和基于遮挡的划分。在随机划分下，任何一种表示都没有始终占优，这表明标准协议可能隐藏了有意义的差异。在主体不重叠评估中，监督关键点显示出统计显著的优势，但性能因主体而异：当解剖标志可见时，它们表现更好，而无监督关键点对遮挡和部分可见性更具鲁棒性，尽管它们在复杂活动中产生更多的假阳性。在基于遮挡的评估中，监督关键点几乎错过了一半的跌倒，而无监督关键点保持了强烈的敏感性，并显著超越了它们。它们的解剖独立性使空间锚点能够适应可见的身体结构，而不是在缺失标志时失败。在带宽限制下，监督定位错误通过时间模型累积，差距进一步扩大。这些发现表明，表示选择应反映预期的视觉条件，并且当身体可见性受到损害时，无监督关键点提供了优势。
cs.CV / 7 / 2607.15401
GS-RealBlur: A Flexible Data Acquisition Framework for Real-World Image Deblurring
GS-RealBlur：一种灵活的真实世界图像去模糊数据获取框架
Abstract
High-quality, large-scale paired data is essential for training learning-based image deblurring models. However, synthetic blurry images generally lack realism, while real-world captured images require complex and inflexible camera systems. In this work, we propose GS-RealBlur, a data acquisition framework for real-world image deblurring, achieving both blur realism and acquisition flexibility. Specifically, we use a handheld camera to capture blurry images, and deploy a gimbal to densely capture sharp images of the same scene. We reconstruct the 3D representation of sharp images and calibrate the camera pose of each blurry frame within this 3D. The image rendered from this 3D according to the pose serves as the sharp counterpart. To better align the rendered image with the blurry image, we introduce a Blur-aware Pose Refinement (BPR) module that refines the pose using appearance consistency and centroid alignment constraints. Leveraging GS-RealBlur, we construct a high-quality and diverse dataset. Extensive experiments demonstrate that a deblurring model trained on our dataset achieves superior generalization performance across various real-world deblurring benchmarks, consistently outperforming models trained on existing synthetic and real-world datasets. The code and dataset will be made publicly available.
Chinese Translation
高质量、大规模的配对数据对于训练基于学习的图像去模糊模型至关重要。然而，合成模糊图像通常缺乏真实感，而真实世界捕获的图像则需要复杂且不灵活的相机系统。在本研究中，我们提出了GS-RealBlur，一种用于真实世界图像去模糊的数据获取框架，既实现了模糊的真实感，又具备了获取的灵活性。具体而言，我们使用手持相机捕获模糊图像，并部署云台密集捕获同一场景的清晰图像。我们重建清晰图像的3D表示，并在该3D空间中校准每个模糊帧的相机姿态。根据姿态从该3D空间渲染的图像作为清晰对应图像。为了更好地将渲染图像与模糊图像对齐，我们引入了一种模糊感知姿态优化（Blur-aware Pose Refinement, BPR）模块，该模块利用外观一致性和质心对齐约束来优化姿态。借助GS-RealBlur，我们构建了一个高质量且多样化的数据集。大量实验表明，在我们的数据集上训练的去模糊模型在各种真实世界去模糊基准测试中表现出色，始终优于在现有合成和真实世界数据集上训练的模型。代码和数据集将公开发布。
cs.CV / 8 / 2607.15410
Explicit Over Implicit: Enhancing CNNs Via Complex Structure Tensor Representations for Periocular Recognition
显式优于隐式：通过复杂结构张量表示增强卷积神经网络在眼周识别中的应用
Abstract
Our study provides evidence that CNNs struggle to extract orientation features effectively. We show that using the Complex Structure Tensor, which contains compact orientation features with certainties, as input to CNNs consistently improves identification accuracy compared to grayscale inputs alone. Experiments also demonstrated that our inputs, provided by mini-complex convnets, combined with reduced CNN sizes, outperformed full-fledged, prevailing CNN architectures. This suggests that the upfront use of orientation features in CNNs, a strategy seen in mammalian vision, not only mitigates their limitations but also enhances their explainability and relevance to thin-clients. Experiments were conducted on publicly available datasets comprising periocular images (Cross-Eyed and PolyU) for biometric identification and verification in both Close-World and Open-World Scenarios using six CNN architectures. Our experiments on the Cross-Eyed and PolyU datasets yield a 5-26% reduction in EER, providing strong empirical evidence that explicit orientation priors mitigate CNN representational limits in Open-World and Close-World scenarios.
Chinese Translation
我们的研究提供了证据，表明卷积神经网络（CNN）在有效提取方向特征方面存在困难。我们展示了使用包含紧凑方向特征及其置信度的复杂结构张量作为CNN输入，相较于仅使用灰度输入，能够持续提高识别准确率。实验还表明，我们通过迷你复杂卷积网络（mini-complex convnets）提供的输入，结合缩小的CNN规模，优于完整的主流CNN架构。这表明，在CNN中前期使用方向特征的策略（这一策略在哺乳动物视觉中常见）不仅缓解了它们的局限性，还增强了它们的可解释性和与薄客户端的相关性。实验在公开可用的数据集上进行，包括用于生物识别识别和验证的眼周图像（Cross-Eyed和PolyU），在封闭世界和开放世界场景中使用六种CNN架构。我们在Cross-Eyed和PolyU数据集上的实验显示，EER降低了5-26%，提供了强有力的实证证据，表明显式方向先验在开放世界和封闭世界场景中缓解了CNN的表征限制。
cs.CV / 9 / 2607.15416
Dataset-Origin Signatures and Shortcut Learning in Screening Mammography AI: A Cross-Dataset Case Study
数据集来源特征与筛查乳腺摄影人工智能中的快捷学习：跨数据集案例研究
Abstract
Reliable AI for screening mammography requires training data representative of the low cancer prevalence and subtle abnormalities found in screening populations. We examined whether supplementing such data with biopsy-confirmed cases from abnormal-enriched external datasets improves performance. Using the Newfoundland and Labrador Breast Screening Dataset (NLBSD) alongside CBIS-DDSM and CMMD, we evaluated an EfficientNet-B5 encoder initialized with Mammo-CLIP weights as a frozen linear probe under consistent preprocessing and patient-level splits. The NLBSD-only model achieved an AUC-ROC of 0.737 (95% CI [0.686, 0.785]). Adding external positive cases reduced performance in every configuration (AUC-ROC = 0.620--0.644; DeLong test, Holm-corrected $p < 0.05$), with degradation increasing as additional sources were introduced. Domain-matched evaluation produced modest gains only when the training and test domains coincided, and no configuration surpassed the NLBSD-only model. As a diagnostic, we reframed the task as predicting each examination's dataset of origin. The datasets were separated almost perfectly despite identical preprocessing, indicating that dataset-specific characteristics strongly influence the learned representation. These findings show that na\"ively pooling abnormal-enriched mammography datasets can introduce domain shift that outweighs the benefit of additional positive cases. Differences in acquisition, intensity mapping, and dataset construction persist after normalization, motivating domain-aware strategies for combining heterogeneous mammography datasets.
Chinese Translation
可靠的筛查乳腺摄影人工智能需要训练数据能够代表筛查人群中低癌症发生率和微妙异常的特征。我们研究了是否通过补充来自异常富集外部数据集的活检确认病例来提高性能。使用纽芬兰和拉布拉多乳腺筛查数据集（NLBSD）与CBIS-DDSM和CMMD，我们评估了一个以Mammo-CLIP权重初始化的EfficientNet-B5编码器，作为在一致预处理和患者级别分割下的冻结线性探针。仅使用NLBSD模型的AUC-ROC达到了0.737（95% CI [0.686, 0.785]）。在每种配置中，添加外部阳性病例均降低了性能（AUC-ROC = 0.620--0.644；DeLong检验，Holm校正$p < 0.05$），且随着额外来源的引入，性能下降加剧。领域匹配评估仅在训练和测试领域重合时产生适度增益，且没有任何配置超越NLBSD-only模型。作为诊断，我们将任务重新框定为预测每次检查的数据集来源。尽管预处理相同，这些数据集几乎完美分离，表明数据集特定特征对学习表示有强烈影响。这些发现表明，简单地合并异常富集的乳腺摄影数据集可能会引入领域转移，从而抵消额外阳性病例的好处。获取、强度映射和数据集构建的差异在归一化后仍然存在，这促使我们采用领域感知策略来结合异质乳腺摄影数据集。
cs.CV / 10 / 2607.15491
Trajectory-aware Cross-view Geo-localization with Sequential Observations
基于轨迹的跨视角地理定位与序列观测
Abstract
Cross-view geo-localization matches ground-level observations against geo-tagged satellite imagery. Recent methods show that sequential queries such as video clips yield richer spatiotemporal cues than single images, yet they overlook a complementary sequential modality: route descriptions -- which capture the same trajectory at a higher level of abstraction and are often the only input available (e.g., a user directing an autonomous vehicle to a pickup point). To bridge this gap, we introduce SeqGeo-VL, a dataset of $\sim$39K video-text-satellite triplets, and TrajLoc, a unified framework capable of processing both video clips and route descriptions. By leveraging both dense visual and abstract linguistic semantics, TrajLoc enables these modalities to mutually reinforce cross-view matching. We further propose TrajMod, a lightweight module that conditions query embeddings on trajectory geometry, yielding spatially-aware representations. Experiments show that TrajLoc achieves substantial gains over state-of-the-art methods on both video and text geo-localization. The project page is available at https://humblegamer.github.io/trajloc/.
Chinese Translation
跨视角地理定位将地面观测与地理标记的卫星图像进行匹配。近期的方法表明，视频片段等序列查询比单幅图像提供了更丰富的时空线索，但它们忽视了一种互补的序列模态：路线描述——这种描述以更高的抽象层次捕捉相同的轨迹，且往往是唯一可用的输入（例如，用户指导自动驾驶车辆到达接送点）。为了解决这一问题，我们引入了SeqGeo-VL，一个包含约39K视频-文本-卫星三元组的数据集，以及TrajLoc，一个能够处理视频片段和路线描述的统一框架。通过利用密集的视觉和抽象的语言语义，TrajLoc使得这两种模态能够相互增强跨视角匹配。我们进一步提出了TrajMod，一个轻量级模块，它根据轨迹几何条件化查询嵌入，从而生成空间感知的表示。实验表明，TrajLoc在视频和文本地理定位上均显著超越了最先进的方法。项目页面可访问 https://humblegamer.github.io/trajloc/。
cs.CV / 11 / 2607.15509
LLM-Driven AutoML for Cross-Lingual Handwritten OCR: Closed-Loop Neural Architecture Search with GPT-5, GPT-4o, and Claude Sonnet 4
基于大型语言模型的跨语言手写光学字符识别自动机器学习：使用GPT-5、GPT-4o和Claude Sonnet 4的闭环神经架构搜索
Abstract
We present a fully automated closed-loop AutoML framework that uses GPT-5, GPT-4o, and Claude Sonnet 4 as autonomous neural architecture designers for cross-lingual handwritten optical character recognition. Each large language model independently generates, trains, evaluates, and iteratively refines neural network architectures using performance feedback from previous trials. The framework is evaluated on Arabic, Persian, and English handwriting datasets through 270 independent experiments. It consistently discovers accurate and computationally efficient models without manual architecture design, domain-specific preprocessing, or hyperparameter tuning. The generated models achieve mean test accuracies above 93 percent, a best accuracy of 98.1 percent, and inference latency between 41 and 44 milliseconds. The results demonstrate that large language models can function as effective AutoML agents for neural architecture search, enabling scalable, script-adaptive, and reproducible handwriting recognition across languages.
Chinese Translation
我们提出了一个完全自动化的闭环自动机器学习框架，该框架使用GPT-5、GPT-4o和Claude Sonnet 4作为自主神经架构设计师，应用于跨语言手写光学字符识别。每个大型语言模型独立生成、训练、评估并迭代优化神经网络架构，利用来自先前实验的性能反馈。该框架在阿拉伯语、波斯语和英语手写数据集上进行了270次独立实验评估。它始终能够发现准确且计算效率高的模型，无需手动架构设计、特定领域的预处理或超参数调整。生成的模型在测试集上的平均准确率超过93%，最佳准确率为98.1%，推理延迟在41到44毫秒之间。结果表明，大型语言模型可以作为有效的自动机器学习代理，进行神经架构搜索，从而实现跨语言的可扩展、脚本自适应和可重复的手写识别。
cs.CV / 12 / 2607.15512
Intentional Electromagnetic Interference Attacks on Facial Recognition
针对面部识别的故意电磁干扰攻击
Abstract
Attacks on general computer vision algorithms are often relegated to the digital domain, with the optimization performed purely in the digital world and then translated to physical mediums for implementation. In the field of biometrics, including facial recognition, physical presentation attacks targeting biometric sensors are dominant and present significant opportunity and risk. This paper highlights a critical vulnerability in the physical-to-digital pipeline of biometric sensors and provides a standardized approach for testing facial recognition system robustness against hardware attacks, going beyond and potentially complementing presentation attacks (as defined in ISO/IEC 30107 standard series). Specifically, in this work we (a) demonstrate that intentional electromagnetic interference is possible to be conducted with commonly accessible radio frequency (RF) equipment, (b) assess the robustness of state-of-the-art face recognition methods against RF-based attacks, and (c) provide a dataset composed of face images captured with and without electromagnetic interference to serve as a new benchmark for testing modern face matchers against RF-sourced interference.
Chinese Translation
对一般计算机视觉算法的攻击通常局限于数字领域，优化过程完全在数字世界中进行，然后转化为物理媒介以实现。在生物识别领域，包括面部识别，针对生物识别传感器的物理呈现攻击占主导地位，并带来了显著的机会和风险。本文强调了生物识别传感器物理到数字管道中的一个关键漏洞，并提供了一种标准化的方法来测试面部识别系统对硬件攻击的鲁棒性，超越并可能补充呈现攻击（如ISO/IEC 30107标准系列中定义的）。具体而言，在本研究中，我们(a) 证明了可以使用常见的射频（RF）设备进行故意电磁干扰，(b) 评估了最先进的面部识别方法对基于RF的攻击的鲁棒性，以及(c) 提供了一个包含带有和不带有电磁干扰的面部图像的数据集，以作为测试现代面部匹配器对抗RF源干扰的新基准。
cs.CV / 13 / 2607.15517
SLAPBench: Benchmarking Multimodal Large Language Models for Four-Finger SLAP Fingerprint Verification
SLAPBench：针对四指SLAP指纹验证的多模态大型语言模型基准测试
Abstract
Four-finger SLAP fingerprints are flat live-scan impressions of the index, middle, ring, and little fingers of one hand, used for identity verification in border control and law enforcement. No benchmark has evaluated whether multimodal large language models (MLLMs) can verify identity from SLAP images. We introduce SLAPBench, the first benchmark for MLLM-based four-finger SLAP fingerprint verification, built from NIST SD302b with 7,832 pairs (176 mated, 7,656 non-mated). We evaluate four open-source MLLMs (InternVL3-8B, Qwen2.5-VL-7B, Qwen3-VL-8B, Gemma-3-12B) and the proprietary Claude Opus 4.8 under zero-shot, task-description, and similarity-scoring prompts. Prompting governs verification behavior. Task-description prompting collapses all four open-source models to near-100% False Accept Rate (FAR), and Gemma-3-12B collapses under zero-shot as well; Claude Opus 4.8 alone resists collapse under both binary prompts, giving the best binary result (FAR = 20.2%). Similarity scoring removes collapse across the open-source models and exposes wide capability gaps: Claude reaches AUC = 0.953 and Gemma-3-12B 0.837, while InternVL3-8B is inverted (AUC = 0.590) and Qwen2.5-VL-7B near random (0.567). Qwen3-VL-8B attains perfect separation (AUC = 1.000), which we treat as a diagnostic rather than as capability: SD302b holds one SLAP capture per finger position, so mated pairs are cross-resolution. A matched-resolution control leaves the perfect score intact, ruling out the resolution shortcut; what cannot be excluded within SD302b is near-duplicate detection, since a mated pair is one capture rendered twice. A fairness probe over gender, race, and age suggests disparity grows as discrimination weakens. SLAPBench establishes the first SLAP-specific MLLM baseline and shows that prompting governs collapse while model capability governs discrimination.
Chinese Translation
四指SLAP指纹是对一只手的食指、中指、无名指和小指的平面活体扫描印象，主要用于边境控制和执法中的身份验证。目前尚无基准评估多模态大型语言模型（MLLMs）是否能够从SLAP图像中验证身份。我们引入SLAPBench，这是首个基于MLLM的四指SLAP指纹验证基准，基于NIST SD302b构建，包含7,832对样本（176对匹配，7,656对不匹配）。我们评估了四个开源MLLM（InternVL3-8B、Qwen2.5-VL-7B、Qwen3-VL-8B、Gemma-3-12B）和专有的Claude Opus 4.8，在零-shot、任务描述和相似性评分提示下进行测试。提示方式决定了验证行为。任务描述提示使所有四个开源模型的误接受率（FAR）接近100%，而Gemma-3-12B在零-shot下也出现崩溃；只有Claude Opus 4.8在这两种二元提示下保持稳定，取得最佳二元结果（FAR = 20.2%）。相似性评分消除了开源模型的崩溃现象，并揭示了显著的能力差距：Claude的AUC达到0.953，Gemma-3-12B为0.837，而InternVL3-8B则呈现倒置（AUC = 0.590），Qwen2.5-VL-7B接近随机（0.567）。Qwen3-VL-8B实现了完美分离（AUC = 1.000），我们将其视为诊断而非能力：SD302b在每个手指位置只包含一个SLAP捕获，因此匹配对是跨分辨率的。匹配分辨率的控制保持了完美分数，排除了分辨率捷径；在SD302b中无法排除的是近重复检测，因为一个匹配对是同一捕获的两次呈现。对性别、种族和年龄的公平性探测表明，随着歧视减弱，差异性增加。SLAPBench建立了首个SLAP特定的MLLM基准，并显示提示方式决定崩溃，而模型能力决定歧视性。
cs.CV / 14 / 2607.15527
Physics-aware Masked Diffusion-based Flood Simulation for Urban Fisheye Disaster Detection
基于物理感知的掩蔽扩散洪水模拟用于城市鱼眼灾害检测
Abstract
Physical simulations that predict the behavior of urban disasters, such as climate-related flooding, play a crucial role in disaster prevention and the development of anomaly detection models. However, the severe shortage of flood data in real-world environments, combined with the inherent distortions of fisheye lens images, which are used for urban surveillance, has made high-precision simulations challenging. To address this, we propose a new physical simulation system PhysFlood that leverages Diffusion Models to synthesize realistic floods from just a single image captured by a fisheye lens. Our system not only enables simulation from a single image, but also features the ability to freely control and generate diverse flood scenarios by manipulating physically meaningful variables, such as water levels. In our evaluation experiments, we conducted a qualitative human study and demonstrated that the simulation images generated by PhysFlood exhibit both acceptable realism and robustness.
Chinese Translation
预测城市灾害（如气候相关洪水）行为的物理模拟在灾害预防和异常检测模型的发展中发挥着至关重要的作用。然而，现实环境中洪水数据的严重短缺，加上用于城市监控的鱼眼镜头图像固有的失真，使得高精度模拟变得具有挑战性。为了解决这一问题，我们提出了一种新的物理模拟系统PhysFlood，该系统利用扩散模型（Diffusion Models）从仅仅一张鱼眼镜头拍摄的图像合成逼真的洪水。我们的系统不仅能够从单张图像进行模拟，还具备通过操控物理上有意义的变量（如水位）自由控制和生成多样化洪水场景的能力。在我们的评估实验中，我们进行了定性的人类研究，证明PhysFlood生成的模拟图像在现实感和鲁棒性方面均表现出可接受的水平。
cs.CV / 15 / 2607.15536
E3DGS: Unified Geometric-Photometric Equivariance for 3D Gaussian Splatting via Color-as-Geometry Embedding
E3DGS：通过颜色作为几何嵌入实现3D高斯点云的统一几何-光度等变性
Abstract
3D Gaussian Splatting (3DGS) captures scenes by coupling explicit geometry (position, covariance) with view-dependent photometry (Spherical Harmonics). However, building $\mathrm{SE}(3)$-equivariant architectures on these primitives presents a fundamental representation bottleneck. Color has been treated as a signal rather than a geometric entity, making it nontrivial to unify symmetry across geometry and appearance as the camera frame changes. While translations are handled by relative coordinates, rotations act heterogeneously across attributes: $\mu\mapsto R\mu$, $\Sigma\mapsto R\Sigma R^\top$, and $f_\ell\mapsto D^\ell(R)f_\ell$. This mismatch complicates strict equivariance, leading existing methods to either discard or flatten SH coefficients, thereby breaking symmetry. We propose a unified solution rooted in representation theory: for SH degrees $\ell\le2$, photometry is algebraically isomorphic to a rank-2 geometric tensor. We prove that the Wigner-$D$ action on these SH coefficients can be exactly reformulated as the conjugation action on $3\times3$ matrices. Leveraging this, we introduce the Unified Matrix Embedding, a lifting that maps all Gaussian attributes into a unified carrier space, $\mathfrak{gl}(3)$. Building on the "Color-as-Geometry" formulation, we present E3DGS, a rigid-body ($\mathrm{SE}(3)$) equivariant architecture that processes 3D Gaussians without Clebsch-Gordan tensor products. Evaluations on object vision and action-conditioned Gaussian world modeling demonstrate that our unified approach yields strong robustness under camera-frame changes and improved data efficiency.
Chinese Translation
3D高斯点云（3DGS）通过将显式几何（位置、协方差）与视角依赖的光度（球谐函数）相结合来捕捉场景。然而，在这些基本元素上构建$ ext{SE}(3)$-等变架构存在根本的表示瓶颈。颜色被视为信号而非几何实体，使得在相机框架变化时统一几何与外观的对称性变得复杂。虽然平移通过相对坐标处理，但旋转在属性间的作用是不均匀的：$
u o R
u$，$oldsymbol{ ext{Σ}} o Roldsymbol{ ext{Σ}} R^ op$，以及$f_ ext{ℓ} o D^ ext{ℓ}(R)f_ ext{ℓ}$。这种不匹配使得严格的等变性变得复杂，导致现有方法要么丢弃，要么压平SH系数，从而破坏了对称性。我们提出了一种基于表示理论的统一解决方案：对于SH度数$ ext{ℓ} ext{≤}2$，光度在代数上同构于一个秩为2的几何张量。我们证明了Wigner-$D$作用在这些SH系数上的确可以被精确地重新表述为对$3 imes3$矩阵的共轭作用。利用这一点，我们引入了统一矩阵嵌入，这是一种将所有高斯属性映射到统一载体空间$ ext{gl}(3)$的提升。基于“颜色作为几何”的表述，我们提出了E3DGS，这是一种刚体（$ ext{SE}(3)$）等变架构，能够处理3D高斯而无需Clebsch-Gordan张量积。在物体视觉和基于动作的高斯世界建模的评估中，我们的统一方法在相机框架变化下展现出强大的鲁棒性和更高的数据效率。
cs.CV / 16 / 2607.15542
ImprovedVBGS: Real-time Continual Variational Bayes Gaussian Splatting
改进的VBGS：实时持续变分贝叶斯高斯点云重建
Abstract
On-the-fly reconstruction is a key requirement for many applications in robotics and autonomous navigation. Variational Bayes Gaussian Splatting (VBGS) enables continual learning without replay buffers using Coordinate Ascent Variational Inference (CAVI), but its per-frame iterations over all observed points make it too slow for real-time use with strict memory and latency requirements. We present ImprovedVBGS, an accelerated framework for on-the-fly continual reconstruction. This is achieved primarily through (i) spatially truncated variational inference, and (ii) improved reassignment that uses forwarding, truncation and eliminates wasteful dynamic recompilation. On the NeRF synthetic dataset, we reduce mean per-frame latency from ~84.0 s to ~0.050 s on an RTX 3070 Ti, a 1680x speed-up while maintaining reconstruction quality.
Chinese Translation
即时重建是机器人技术和自主导航中许多应用的关键需求。变分贝叶斯高斯点云重建（Variational Bayes Gaussian Splatting, VBGS）通过使用坐标上升变分推理（Coordinate Ascent Variational Inference, CAVI）实现了无需重放缓冲区的持续学习，但其对所有观察点的逐帧迭代使其在具有严格内存和延迟要求的实时应用中速度过慢。我们提出了改进的VBGS（ImprovedVBGS），这是一个加速的即时持续重建框架。该框架主要通过（i）空间截断变分推理和（ii）改进的重新分配方法来实现，后者利用了前向传播、截断并消除了不必要的动态重编译。在NeRF合成数据集上，我们将每帧的平均延迟从约84.0秒降低到约0.050秒，在RTX 3070 Ti上实现了1680倍的加速，同时保持了重建质量。
cs.CV / 17 / 2607.15556
When Can Test-Time Adaptation Help Zero-Shot CT Vision-Language Models?
何时测试时适应能帮助零样本CT视觉语言模型？
Abstract
3D CT vision-language models (VLMs) classify abnormalities from text prompts in a zero-shot manner, enabling cross-institution deployment where labels are scarce and clinical tasks shift faster than supervised models can be retrained. A real CT scan, however, typically contains several co-occurring abnormalities, and the reliability of zero-shot multi-label prediction under distribution shift remains poorly understood. Test-time adaptation (TTA) updates a model on unlabeled target scans without source data or target annotations, yet existing TTA methods target multi-class softmax prediction on natural images or 2D medical segmentation, and none addresses unsupervised multi-label adaptation for zero-shot 3D CT VLMs. We study when TTA helps zero-shot 3D CT VLMs. A controlled diagnostic analysis shows that TTA is conditional: the volumetric input must preserve the encoder's depth structure, and the base representation must transfer to the target cohort, with depth reduction alone lowering internal AUROC by more than 0.12. We then focus on the regime where the base model already separates present from absent abnormalities. We introduce CARVE (Cardinality-Aware Retained-View Entropy), the first TTA method for this setting. CARVE estimates a sample-specific positive-label cardinality $\hat{k}$, optimizes a top-$\hat{k}$ objective to preserve co-occurring abnormalities, and performs memory-efficient multi-view adaptation by scoring weak 3D views without gradients before updating on a retained subset. Across contrastive CT-CLIP and anatomy-aware fVLM, CARVE provides the most consistent improvements across multi-label, three-class, and binary CT tasks when the base model is already discriminative. These results establish multi-label TTA for zero-shot 3D CT VLMs as a distinct problem and CARVE as a cardinality-aware solution.
Chinese Translation
3D CT视觉语言模型（VLMs）以零样本方式根据文本提示对异常进行分类，从而实现跨机构部署，在标签稀缺且临床任务变化速度超过监督模型重训练速度的情况下尤为重要。然而，真实的CT扫描通常包含多个共现的异常，而在分布变化下零样本多标签预测的可靠性仍然不甚明了。测试时适应（TTA）在没有源数据或目标注释的情况下更新模型，但现有的TTA方法主要针对自然图像或2D医学分割中的多类softmax预测，尚无针对零样本3D CT VLMs的无监督多标签适应方法。我们研究了TTA何时能帮助零样本3D CT VLMs。受控的诊断分析表明，TTA是有条件的：体积输入必须保留编码器的深度结构，基础表示必须能够转移到目标队列，仅深度减少就会使内部AUROC降低超过0.12。接着，我们关注于基础模型已经能够区分存在与缺失异常的情形。我们引入了CARVE（基数感知保留视图熵），这是该设置下的首个TTA方法。CARVE估计样本特定的正标签基数$ ext{hat{k}}$，优化一个top-$ ext{hat{k}}$目标以保留共现异常，并通过在更新保留子集之前对弱3D视图进行无梯度评分来执行内存高效的多视图适应。在对比CT-CLIP和解剖感知fVLM的实验中，当基础模型已经具有区分能力时，CARVE在多标签、三类和二元CT任务中提供了最一致的改进。这些结果将零样本3D CT VLMs的多标签TTA确立为一个独特的问题，并将CARVE视为一种基数感知的解决方案。
cs.CV / 18 / 2607.15563
Are All Tokens Necessary for Visual Place Recognition? An Empirical Study of Token Reduction for Efficient Inference
所有视觉标记在视觉位置识别中都是必要的吗？一种关于标记减少以实现高效推理的实证研究
Abstract
Recent visual place recognition (VPR) methods based on vision transformers, particularly foundation models, have achieved remarkable recognition performance. However, these models process all visual tokens throughout the entire network, resulting in substantial computational overhead, which hinders their deployment in real-time and resource-constrained scenarios. A natural question thus arises: are all visual tokens necessary for VPR? To answer this question, we present the first systematic benchmark of token reduction for efficient visual place recognition. Our benchmark comprehensively evaluates representative token pruning, token merging, and hybrid pruning-merging methods across multiple state-of-the-art VPR models and diverse benchmark datasets covering urban, suburban, and natural environments. We further investigate token reduction from multiple perspectives, including recognition performance under different reduction configurations, computational complexity, inference speed, qualitative visualization, and deployment efficiency on edge devices. Through extensive experiments and in-depth analysis, our benchmark reveals multiple important characteristics of token reduction in VPR and provides several practical insights into the trade-offs between accuracy and inference efficiency. For example, token reduction can reduce computational cost by up to 29\% and improve throughput by up to 44\%, while incurring less than 1\% degradation in recognition accuracy. Overall, this work establishes a comprehensive foundation for future research on token-efficient VPR and efficient visual retrieval systems. Our codes and models will be available at https://github.com/Tong-Jin01/TokenReduction4VPR
Chinese Translation
基于视觉变换器的近期视觉位置识别（VPR）方法，特别是基础模型，已取得显著的识别性能。然而，这些模型在整个网络中处理所有视觉标记，导致了巨大的计算开销，从而阻碍了它们在实时和资源受限场景中的部署。因此，一个自然的问题随之而来：所有视觉标记在VPR中都是必要的吗？为了解答这个问题，我们提出了第一个系统的标记减少基准，以实现高效的视觉位置识别。我们的基准全面评估了多种代表性的标记修剪、标记合并和混合修剪-合并方法，涵盖了多个最先进的VPR模型和涵盖城市、郊区和自然环境的多样基准数据集。我们进一步从多个角度研究标记减少，包括在不同减少配置下的识别性能、计算复杂度、推理速度、定性可视化以及在边缘设备上的部署效率。通过广泛的实验和深入的分析，我们的基准揭示了VPR中标记减少的多个重要特性，并提供了关于准确性与推理效率之间权衡的若干实用见解。例如，标记减少可以将计算成本降低多达29%，并将吞吐量提高多达44%，同时识别准确性下降不足1%。总体而言，这项工作为未来关于标记高效VPR和高效视觉检索系统的研究奠定了全面的基础。我们的代码和模型将发布在 https://github.com/Tong-Jin01/TokenReduction4VPR
cs.CV / 19 / 2607.15565
Ask Twice, Look Twice: Prompt Echoing Resolves the Question-First Paradox in Vision-Language Models
问两次，看两次：提示回声解决视觉-语言模型中的问题优先悖论
Abstract
Where should the question go in a vision-language model (VLM) prompt: before the image or after it? Intuition says before: knowing what is asked should tell the model where to look. Yet across visual question answering benchmarks, question-first prompting consistently underperforms the image-first ordering recommended for frontier VLMs, a phenomenon we term the question-first paradox. We trace the paradox to a conflict between two stages of VLM computation. Logit-lens and attention probes show the intuition is half right: a question placed before the image genuinely steers perception, moving image patch representations toward question-relevant concepts. The failure lies downstream. Stranded behind hundreds of image tokens, the question is barely attended by the answer token, which instead commits to image-driven (often wrong) answers; a causal attention knockout confirms that the answer reads the question only when the question follows the image. The diagnosis yields a training-free fix: question echoing, restating the question on both sides of the image so that one copy steers perception while the other is read out at answer time. The same division of labor appears in a fifty-year-old finding on human ``adjunct questions'', where repeating a question before and after a passage aids comprehension more than either position alone. Echoing the image as well brings further gains, restoring the whole-image view a causal decoder otherwise loses. The paradox holds across five open VLMs, costing up to 17.5 group-accuracy points. Echoed prompts close it and surpass the best single-pass ordering on NaturalBench, POPE, Winoground, and open-ended VQAv2, by up to 19 Winoground group-accuracy points, with no training, fine-tuning, or architecture change. The paradox reveals a trade-off between steering perception and preserving question access; echoing resolves it through prompt design alone.
Chinese Translation
在视觉-语言模型（VLM）提示中，问题应该放在图像之前还是之后？直觉告诉我们应该放在之前：知道所问内容应能指引模型的注意力。然而，在视觉问答基准测试中，问题优先的提示方式始终表现不如推荐的图像优先顺序，这一现象我们称之为问题优先悖论。我们追溯这一悖论的根源在于VLM计算的两个阶段之间的冲突。Logit-lens和注意力探测表明，直觉是部分正确的：放在图像之前的问题确实能够引导感知，将图像补丁表示向与问题相关的概念移动。但失败发生在后续阶段。在数百个图像标记之后，问题几乎没有被答案标记关注，答案标记则倾向于做出基于图像的（通常是错误的）回答；因果注意力的消失确认了答案只有在问题位于图像之后时才会读取该问题。这个诊断提供了一种无训练的解决方案：问题回声，即在图像的两侧重述问题，使得一份副本引导感知，而另一份在回答时被读取。相同的分工出现在五十年前关于人类“附加问题”的发现中，在段落之前和之后重复问题比单独任一位置更有助于理解。回声图像也带来了进一步的收益，恢复了因果解码器通常会丢失的全图视图。该悖论在五个开放的VLM中均存在，损失高达17.5个组准确率点。回声提示弥补了这一差距，并在NaturalBench、POPE、Winoground和开放式VQAv2上超越了最佳单次通过顺序，提升了多达19个Winoground组准确率点，且无需训练、微调或架构更改。该悖论揭示了引导感知与保持问题访问之间的权衡；回声通过提示设计单独解决了这一问题。
cs.CV / 20 / 2607.15600
Geometric Distillation from Rectified Stereo: Leveraging Epipolar Cues for Monocular Depth
来自校正立体图的几何蒸馏：利用极线线索进行单目深度估计
Abstract
Monocular depth foundation models have demonstrated remarkable generalization capabilities across diverse environments. However, they continue to struggle with metric depth estimation in diverse environments. This limitation stems from the inherent scale ambiguity of single-view inference, leading to misaligned scale predictions even when the relative geometry is accurate. Conversely, recent multi-view foundation models leverage cross-view cues to learn robust scene-level geometry and consistent scale. Yet, these benefits typically vanish during single-image inference, as the absence of explicit geometric constraints causes performance to degrade. To bridge this gap, we propose a novel framework that transfers the scale-aware geometric priors of multi-view models into monocular depth foundation models. Specifically, we introduce an Epipolar Distillation (EpiDistill), an approach utilizing Rectified Stereo Tokens, which enables the single-view prediction model to retain epipolar attention patterns and maintain geometric consistency without requiring multi-view inputs at inference. Experimental results demonstrate that our method significantly improves zero-shot metric depth estimation, particularly on challenging datasets like ETH3D and DIODE where scale alignment is critical. Furthermore, our approach is model-agnostic, consistently boosting the performance of state-of-the-art ViT-based models, including UniDepthV2 and DepthPro.
Chinese Translation
单目深度基础模型在多样化环境中展示了显著的泛化能力。然而，它们在多样化环境中的度量深度估计仍然面临挑战。这一限制源于单视图推断固有的尺度模糊性，即使相对几何关系准确，尺度预测也会出现不对齐。相反，近期的多视图基础模型利用跨视图线索来学习稳健的场景级几何和一致的尺度。然而，这些优势在单图像推断过程中通常会消失，因为缺乏明确的几何约束导致性能下降。为了解决这一问题，我们提出了一种新颖的框架，将多视图模型的尺度感知几何先验转移到单目深度基础模型中。具体而言，我们引入了极线蒸馏（Epipolar Distillation, EpiDistill）方法，该方法利用校正立体标记，使单视图预测模型能够保留极线注意模式，并在推断时保持几何一致性，而无需多视图输入。实验结果表明，我们的方法显著提高了零-shot度量深度估计的性能，特别是在ETH3D和DIODE等具有挑战性的数据集上，其中尺度对齐至关重要。此外，我们的方法是模型无关的，持续提升了以ViT为基础的最先进模型的性能，包括UniDepthV2和DepthPro。
cs.CV / 21 / 2607.15604
WREN: Low Light Image Enhancement Using Retinex theory-based Double U-Net-like Structures
WREN：基于Retinex理论的双U-Net结构低光照图像增强
Abstract
This paper proposes a neural network for low light image enhancement (LLIE) based on retinex theory to make LLIE robust for various dynamic range scenes. The retinex theory is an image formulation model inspired by a human color perception hypothesis, where a low light image is decomposed into intrinsic color context (i.e., reflectance map) and scene-dependent illumination (i.e., illumination map). Due to non-uniqueness of its decomposition, existing retinex-based LLIE methods often fail to achieve stable decomposition, which lead to over-enhancement. Typically, they are sensitive to the dynamic ranges that vary in different lighting conditions. To tackle this issue, we propose WREN: An LLIE neural network with double U-Net-like structures. WREN consists of two U-Net-like sub-networks. The first network has one encoder and two decoders that decompose an input image into the reflectance and illumination maps. The second network with a customized Transformer block between an encoder and a decoder only enhances the illumination map obtained from the first network: This completely follows the assumption of the retinex theory. Finally, the enhanced illumination map is recombined with the reflectance map. The network is trained end-to-end with a scale-invariant loss function, which gives robustness against the illumination scaling. Numerical results show that our method achieves the state-of-the-art performance across multiple datasets. Our code is available online.
Chinese Translation
本文提出了一种基于Retinex理论的低光照图像增强（LLIE）神经网络，以使LLIE在各种动态范围场景中具有鲁棒性。Retinex理论是一种图像表述模型，灵感来源于人类色彩感知假设，其中低光照图像被分解为内在颜色上下文（即反射率图）和场景依赖的照明（即照明图）。由于其分解的非唯一性，现有的基于Retinex的LLIE方法往往无法实现稳定的分解，导致过度增强。通常，它们对不同照明条件下变化的动态范围敏感。为了解决这个问题，我们提出了WREN：一种具有双U-Net结构的LLIE神经网络。WREN由两个U-Net样的子网络组成。第一个网络具有一个编码器和两个解码器，将输入图像分解为反射率图和照明图。第二个网络在编码器和解码器之间加入了定制的Transformer块，仅增强来自第一个网络的照明图：这完全遵循了Retinex理论的假设。最后，增强后的照明图与反射率图重新组合。该网络采用端到端训练，并使用尺度不变损失函数，使其对照明缩放具有鲁棒性。数值结果表明，我们的方法在多个数据集上达到了最先进的性能。我们的代码已在线发布。
cs.CV / 22 / 2607.15605
Benchmarking MRI Representations for Deep Learning-Based Focal Cortical Dysplasia Segmentation
基于深度学习的局灶性皮质发育不良分割的MRI表示基准测试
Abstract
Focal cortical dysplasia (FCD) is one of the leading structural causes of drug-resistant focal epilepsy, yet its subtle and heterogeneous imaging characteristics make accurate identification and delineation challenging on conventional magnetic resonance imaging (MRI). Although T1-weighted (T1w) and fluid-attenuated inversion recovery (FLAIR) images are routinely acquired for presurgical evaluation, the contribution of different MRI representations to deep learning-based FCD segmentation remains poorly understood. In this study, we present a systematic benchmark of MRI representations for automated FCD segmentation using the nnU-Net framework. A publicly available presurgical MRI dataset comprising 85 FCD subjects and 25 healthy controls was used to evaluate eight input configurations, including conventional MRI contrasts (T1w and FLAIR), ratio-derived representations, and their multimodal combinations. To isolate the effect of MRI representation, all experiments employed identical preprocessing, network architecture, optimization strategy, and five-fold cross-validation. Among the evaluated single-modality representations, FLAIR achieved the strongest overall performance, whereas ratio-derived representations alone were insufficient for reliable identification of subtle FCD. Incorporating ratio-derived representations with conventional T1w and FLAIR images consistently improved lesion delineation, with the four-channel multimodal configuration achieving the highest overall Dice score (0.376), representing a 5.0% relative improvement over the conventional T1w+FLAIR representation. These findings demonstrate that MRI representation design is an important yet underexplored component of deep learning-based FCD segmentation and should be optimized alongside network architecture.
Chinese Translation
局灶性皮质发育不良（FCD）是药物难治性局灶性癫痫的主要结构性原因之一，但其微妙且异质的影像特征使得在常规磁共振成像（MRI）中准确识别和描绘变得具有挑战性。尽管在术前评估中常规获取T1加权（T1w）和液体衰减反转恢复（FLAIR）图像，但不同MRI表示对基于深度学习的FCD分割的贡献仍然不甚了解。在本研究中，我们使用nnU-Net框架对自动化FCD分割的MRI表示进行了系统的基准测试。我们使用一个公开可用的术前MRI数据集，该数据集包含85名FCD患者和25名健康对照，以评估八种输入配置，包括常规MRI对比（T1w和FLAIR）、比率衍生表示及其多模态组合。为了隔离MRI表示的影响，所有实验均采用相同的预处理、网络架构、优化策略和五折交叉验证。在评估的单模态表示中，FLAIR表现出最强的整体性能，而仅使用比率衍生表示不足以可靠地识别微妙的FCD。将比率衍生表示与常规的T1w和FLAIR图像结合，持续改善了病灶的描绘，其中四通道多模态配置达到了最高的整体Dice系数（0.376），比常规的T1w+FLAIR表示提高了5.0%的相对改善。这些发现表明，MRI表示设计是基于深度学习的FCD分割中一个重要但尚未深入探讨的组成部分，应与网络架构一起进行优化。
cs.CV / 23 / 2607.15615
Region-Grounded Vision-Language Learning for Detection-Guided Mammographic Lesion Classification
基于区域的视觉-语言学习用于检测引导的乳腺X光病变分类
Abstract
Vision-language models trained with contrastive objectives have shown promise in medical image analysis. However, conventional global image-text alignment is ill-suited for mammography, where diagnostically relevant lesions are spatially localized and occupy only a small fraction of the image. Subtle morphological cues critical for malignancy assessment can be diluted when representations are learned at the whole-image level. In this work, we propose a novel region-grounded vision-language learning method for detection-guided mammographic lesion classification. The method mirrors radiologists' diagnostic paradigm. First, a region-text contrastive pretraining stage aligns lesion-specific features with structured clinical descriptors derived from radiology metadata. To mitigate semantic collapse and background bias in low-vocabulary settings, we introduce a multi-component objective incorporating positive alignment, fine-grained semantic hard negatives, and background suppression. Second, an auxiliary lesion detection head is jointly optimized with contrastive classification to preserve spatial sensitivity and enable localization-aware malignancy classification. Extensive experiments on two independent datasets, CBIS-DDSM and VinDr-Mammo, show superior performance of our method compared to related methods under in-domain, cross-dataset, and transfer learning settings.
Chinese Translation
使用对比目标训练的视觉-语言模型在医学图像分析中显示出良好的前景。然而，传统的全局图像-文本对齐不适合乳腺X光检查，因为诊断相关的病变是空间局部化的，并且仅占图像的一小部分。在全图级别学习表示时，评估恶性肿瘤所需的细微形态线索可能会被稀释。在本研究中，我们提出了一种新颖的基于区域的视觉-语言学习方法，用于检测引导的乳腺X光病变分类。该方法反映了放射科医师的诊断范式。首先，区域-文本对比预训练阶段将病变特征与来自放射学元数据的结构化临床描述符对齐。为了减轻低词汇环境中的语义崩溃和背景偏差，我们引入了一种多组件目标，结合了正对齐、细粒度语义硬负样本和背景抑制。其次，辅助病变检测头与对比分类共同优化，以保持空间敏感性并实现定位感知的恶性肿瘤分类。在两个独立数据集CBIS-DDSM和VinDr-Mammo上的广泛实验表明，我们的方法在领域内、跨数据集和迁移学习设置下的性能优于相关方法。
cs.CV / 24 / 2607.15619
StructGen: Disambiguating Multi-Reference Image Generation via Structured Context Modeling
StructGen：通过结构化上下文建模消歧义多参考图像生成
Abstract
Multi-reference image generation aims to synthesize images by integrating attributes from multiple reference images under textual instructions. As the number of references increases, the task necessitates complex semantic comprehension, such as correctly associating attributes with the intended subjects and planing out coherent spatial arrangement between subjects and their environments. Existing approaches, which rely solely on natural language instruction, often fail to capture these complex intentions precisely, leading to semantic misalignment and inconsistent generation. We identify two key factors behind these limitations: natural language instructions are often verbose and ambiguous, and high-quality multi-reference data is scarce. To address these issues, we propose StructGen, which employs a structured, dictionary-like format to encode multiple reference images, thereby enabling explicit and unambiguous specification of generation intentions. To support this design, we construct a structured dataset based on high-quality real images and develop a corresponding training framework, along with a dedicated benchmark for challenging multi-reference scenarios. Extensive experiments on both public benchmarks and our proposed benchmark demonstrate that StructGen consistently outperforms existing methods on both semantic alignment and detailed reference-generation consistency, especially under complex instructions with multiple references. The code is available at \href{https://jianingpeng0382.github.io/StructGen/}{https://jianingpeng0382.github.io/StructGen/}.
Chinese Translation
多参考图像生成旨在通过整合来自多个参考图像的属性，在文本指令下合成图像。随着参考图像数量的增加，该任务需要复杂的语义理解，例如正确地将属性与预期对象关联，并规划对象与其环境之间的连贯空间布局。现有方法仅依赖自然语言指令，往往无法精确捕捉这些复杂意图，导致语义不一致和生成结果不稳定。我们识别出这些局限性的两个关键因素：自然语言指令通常冗长且模糊，而高质量的多参考数据稀缺。为了解决这些问题，我们提出了StructGen，它采用结构化的字典式格式对多个参考图像进行编码，从而能够明确且无歧义地指定生成意图。为了支持这一设计，我们基于高质量真实图像构建了一个结构化数据集，并开发了相应的训练框架，以及一个专门针对挑战性多参考场景的基准测试。针对公共基准和我们提出的基准的广泛实验表明，StructGen在语义对齐和详细参考生成一致性方面始终优于现有方法，尤其是在具有多个参考的复杂指令下。代码可在 exttt{https://jianingpeng0382.github.io/StructGen/} 获取。
cs.CV / 25 / 2607.15639
BCG-Former: Toward Pareto-Efficient Hyperspectral Image Classification via Band-Contextual Gating
BCG-Former：通过带上下文门控实现帕累托有效的高光谱图像分类
Abstract
Hyperspectral image (HSI) classification systems are increasingly deployed on platforms with strict computational budgets, such as UAVs and small spaceborne sensors. In these settings, accuracy alone is not enough; the model must also run within tight latency and memory constraints. Most recent HSI classifiers, however, focus on accuracy and pay relatively little attention to these constraints. We propose BCG-Former, a lightweight CNN-Transformer hybrid that targets this trade-off. The model introduces three innovations: (1) Band-Contextual Gating (BCG) for adaptive spectral recalibration using local inter-band context and learnable temperature sharpening, (2) a spectral summary token that bridges spectral and spatial features, and (3) single-pass Band-RoPE combined with linear attention for efficient joint representation learning. Evaluated on classical airborne (Pavia University, Salinas, Indian Pines, Houston 2013/2018) and UAV-borne benchmark datasets (WHU-Hi-LongKou, HongHu, and HanChuan), BCG-Former achieves over-all accuracy ranging from 91.51% on Houston 2018 to 99.49% on Houston 2013, while maintaining sub-millisecond inference latency (0.91-0.95ms) and using only 0.10-0.23M parameters. Across all eight benchmarks, BCG-Former consistently resides on or near the Pareto frontier of accuracy versus latency, outperforming or matching recent CNN-, Transformer-, and Mamba-based methods at a fraction of their computational cost. Ablation studies confirm that all three components are complementary, with BCG providing the largest individual contribution. These results establish BCG-Former as a strong accuracy-efficiency Pareto candidate for real-time and large-scale remote sensing applications.
Chinese Translation
高光谱图像（HSI）分类系统越来越多地部署在计算预算严格的平台上，如无人机（UAV）和小型空间传感器。在这些环境中，单靠准确性是不够的；模型还必须在严格的延迟和内存限制内运行。然而，最近的大多数HSI分类器主要关注准确性，对这些限制关注较少。我们提出了BCG-Former，这是一种轻量级的CNN-Transformer混合模型，旨在解决这一权衡。该模型引入了三项创新：（1）带上下文门控（Band-Contextual Gating, BCG），用于利用局部带间上下文和可学习的温度锐化进行自适应光谱重校准；（2）一种光谱摘要标记，桥接光谱特征和空间特征；（3）单次传递的带-旋转位置编码（Band-RoPE）结合线性注意力，用于高效的联合表示学习。在经典的航空（Pavia University, Salinas, Indian Pines, Houston 2013/2018）和无人机基准数据集（WHU-Hi-LongKou, HongHu, 和 HanChuan）上进行评估，BCG-Former的整体准确率从Houston 2018的91.51%到Houston 2013的99.49%不等，同时保持亚毫秒的推理延迟（0.91-0.95毫秒），并仅使用0.10-0.23M的参数。在所有八个基准测试中，BCG-Former始终位于或接近准确性与延迟的帕累托前沿，超越或匹配最近的基于CNN、Transformer和Mamba的方法，且计算成本仅为其一小部分。消融研究证实这三部分是互补的，其中BCG提供了最大的单独贡献。这些结果确立了BCG-Former作为实时和大规模遥感应用中强大的准确性-效率帕累托候选者。
cs.CV / 26 / 2607.15650
DiTango: Cost-Effective Parallel Diffusion Generation with Selective Attention State Reuse
DiTango：具有选择性注意状态重用的成本效益并行扩散生成
Abstract
Recent advances in AI-generated content have driven widespread adoption of Diffusion Transformers (DiTs) for high-resolution, long-duration content generation. While parallelization techniques accelerate diffusion inference, they face significant scalability challenges due to excessive communication overhead in multi-node environments. We observe that sequence partitions in Context Parallelism (CP) exhibit distinct heterogeneity: spatially proximate partitions contribute more significantly to attention computation results. By mapping this heterogeneous pattern to hierarchical communication topology, we can access high-contribution partitions with reduced communication cost. This insight motivates our novel selective attention state mechanism that strategically balances partial attention computation and historical result reuse across denoising steps. We present DiTango, an efficient parallel framework for DiT generation. DiTango features an anchor-guided state selection planner that optimizes computation-reuse decisions for each partition, complemented by a runtime that orchestrates efficient state-centric operations. This design achieves superior system efficiency while preserving generation quality. Experimental evaluation on popular diffusion models demonstrates that DiTango achieves up to 1.9x end-to-end and 3.2x attention speedup with near-linear scaling in multi-node settings, while maintaining generation quality comparable to state-of-the-art approaches.
Chinese Translation
近期人工智能生成内容的进展推动了扩散变换器（Diffusion Transformers, DiTs）在高分辨率、长时长内容生成中的广泛应用。虽然并行化技术加速了扩散推理，但由于多节点环境中的通信开销过大，它们面临显著的可扩展性挑战。我们观察到上下文并行（Context Parallelism, CP）中的序列划分表现出明显的异质性：空间上相近的划分对注意力计算结果的贡献更为显著。通过将这种异质模式映射到分层通信拓扑，我们可以以降低通信成本的方式访问高贡献划分。这一洞察促使我们提出了一种新颖的选择性注意状态机制，战略性地平衡去噪步骤中的部分注意力计算和历史结果重用。我们提出了DiTango，一个高效的DiT生成并行框架。DiTango具有一个锚点引导的状态选择规划器，优化每个划分的计算重用决策，并配备一个协调高效状态中心操作的运行时。该设计在保持生成质量的同时实现了卓越的系统效率。在流行的扩散模型上的实验评估表明，DiTango在多节点环境中实现了高达1.9倍的端到端和3.2倍的注意力加速，且具有近线性的可扩展性，同时保持与最先进方法相当的生成质量。
cs.CV / 27 / 2607.15652
CSS-BA: Gate-Guided Column Space Search for Bundle Adjustment
CSS-BA：基于门控的列空间搜索用于束调整
Abstract
Bundle adjustment (BA) remains a critical refinement module for image-based 3D reconstruction and continues to improve geometric accuracy even in learning-based pipelines. However, in low-parallax and near-rotational regimes, classical Schur-based Levenberg--Marquardt (LM) often becomes ill-conditioned and yields unreliable pose and calibration estimates. We propose Gate-Guided CSS-BA, a solver-side modification of Schur-LM that preserves the classical BA objective and trust-region framework while constraining each update to a geometrically informed low-dimensional subspace. By integrating Column Space Search (CSS) with geometry-aware gating, the method stabilizes the Schur-LM update without altering the estimation problem. In contrast to keyframe or state-selection approaches, all camera and point parameters remain in the optimization problem; only the update direction is restricted. The method serves as a drop-in replacement for existing BA pipelines. Experiments on both generic and challenging weak-geometry scenarios show more stable optimization, improved relative pose accuracy, and competitive calibration behavior while maintaining reprojection quality.
Chinese Translation
束调整（BA）仍然是基于图像的三维重建中的一个关键优化模块，并且在基于学习的流程中继续提高几何精度。然而，在低视差和近旋转的情况下，经典的基于舒尔（Schur）的列文伯格-马夸特（Levenberg-Marquardt, LM）方法往往变得病态，并产生不可靠的姿态和标定估计。我们提出了门控引导的CSS-BA，这是对舒尔-LM的求解器侧修改，保留了经典BA目标和信任域框架，同时将每次更新限制在几何信息驱动的低维子空间内。通过将列空间搜索（Column Space Search, CSS）与几何感知门控相结合，该方法在不改变估计问题的情况下稳定了舒尔-LM更新。与关键帧或状态选择方法相比，所有相机和点参数仍然保留在优化问题中；仅限制更新方向。该方法可以作为现有BA流程的直接替代。对通用和具有挑战性的弱几何场景的实验表明，该方法实现了更稳定的优化、提高的相对姿态精度和竞争力的标定表现，同时保持了重投影质量。
cs.CV / 28 / 2607.15661
Model Merging for Medical LVLMs: A Benchmark and a Winner-Take-All Approach
医学领域大规模视觉语言模型的模型合并：基准测试与赢家通吃方法
Abstract
Large vision-language models (LVLMs) can be adapted to specialized medical imaging tasks via parameter-efficient fine-tuning approaches such as low-rank adaptation (LoRA), leading to a growing ecosystem of expert models tailored to specific imaging modalities and clinical scenarios. However, deploying multiple expert LVLMs in practice incurs substantial computational and operational overhead. Model merging provides a promising solution by consolidating multiple experts into a single model without retraining, yet it remains largely unexplored in the medical domain. In this work, we present the first systematic study of model merging for medical LVLMs. We introduce MergeMedBench, a comprehensive benchmark spanning eight imaging modalities and diverse clinical task types, comprising 16 LoRA fine-tuned models built upon two mainstream architectures. We conduct an extensive evaluation of existing merging methods and further propose winner-take-all, a simple and hyperparameter-free approach that retains only the most dominant parameters across expert models. By preserving the critical parameters that govern model behavior and discarding weaker ones, our method avoids the information dilution inherent in averaging- or alignment-based strategies. Despite its simplicity, winner-take-all consistently outperforms existing approaches, offering both a new perspective on LoRA merging and a strong practical baseline for future research.
Chinese Translation
大型视觉语言模型（LVLMs）可以通过低秩适应（LoRA）等参数高效微调方法，适应于专业的医学影像任务，从而形成一个日益增长的专家模型生态系统，专门针对特定的影像模态和临床场景。然而，在实践中部署多个专家LVLM会产生大量的计算和操作开销。模型合并提供了一种有前景的解决方案，通过将多个专家整合为一个单一模型而无需重新训练，但在医学领域仍然基本未被探索。在本研究中，我们首次系统性地研究了医学LVLM的模型合并。我们引入了MergeMedBench，这是一个涵盖八种影像模态和多种临床任务类型的综合基准，包括基于两种主流架构构建的16个LoRA微调模型。我们对现有的合并方法进行了广泛评估，并进一步提出了赢家通吃（winner-take-all）方法，这是一种简单且无超参数的方法，仅保留专家模型中最具主导性的参数。通过保留决定模型行为的关键参数并丢弃较弱的参数，我们的方法避免了基于平均或对齐策略的固有信息稀释。尽管其简单性，赢家通吃方法始终优于现有方法，为LoRA合并提供了新的视角，并为未来研究提供了强有力的实践基准。
cs.CV / 29 / 2607.15667
PE-Field 4D: Video Generation Models as Canvas
PE-Field 4D：视频生成模型作为画布
Abstract
Diffusion Transformers have recently achieved strong performance in video generation, yet controlling scene geometry under viewpoint changes and camera motion remains challenging. In this work, we revisit the role of positional encoding in video diffusion transformers and show that it provides a useful spatial bias for geometry-aware control. Specifically, if reference tokens are encoded according to their projected locations in the target view, the denoising model is encouraged to retrieve content from position aligned regions of the input video. Building on this observation, we introduce a geometry-aware cross-attention mechanism that enables target video latent tokens to attend to structured context tokens derived from reference images or frames. To establish correspondence between the reference content and the target camera trajectory, we equip the context tokens with a projected positional encoding scheme that combines target-view 2D reprojection with depth-aware disambiguation. At the same time, we preserve the original spatiotemporal positional encoding of the generated video latent, allowing geometric guidance to be injected while maintaining consistency with the video model's native latent structure. The resulting framework provides a simple and effective approach for controllable video generation. It improves spatial controllability in viewpoint-dependent editing tasks, including camera re-trajectory, novel-view video synthesis, and geometry-aware video editing, while preserving the generative prior of the underlying video diffusion model. The code is available at: https://github.com/MTLab/PE-Field.
Chinese Translation
扩散变换器（Diffusion Transformers）最近在视频生成方面取得了强劲的表现，但在视角变化和相机运动下控制场景几何形状仍然具有挑战性。在本研究中，我们重新审视了位置编码在视频扩散变换器中的作用，并展示了它为几何感知控制提供了有用的空间偏差。具体而言，如果参考标记根据其在目标视图中的投影位置进行编码，去噪模型将被鼓励从输入视频的空间对齐区域中检索内容。在此观察的基础上，我们引入了一种几何感知的交叉注意机制，使目标视频潜在标记能够关注来自参考图像或帧的结构化上下文标记。为了在参考内容与目标相机轨迹之间建立对应关系，我们为上下文标记配备了一种投影位置编码方案，该方案结合了目标视图的二维重投影和深度感知消歧。在此过程中，我们保留了生成视频潜在的原始时空位置编码，从而在保持与视频模型的原生潜在结构一致性的同时，注入几何指导。所提出的框架为可控视频生成提供了一种简单有效的方法。它改善了在视角依赖编辑任务中的空间可控性，包括相机重轨迹、新视角视频合成和几何感知视频编辑，同时保留了基础视频扩散模型的生成先验。代码可在以下链接获取：https://github.com/MTLab/PE-Field。
cs.CV / 30 / 2607.15689
Efficient Frame Selection for Long Videos at Test Time with Attention-Based MLLM Selectors
基于注意力的多模态大语言模型选择器在测试时对长视频的高效帧选择
Abstract
Understanding long videos with multimodal large language models (MLLMs) requires selecting a compact set of frames from thousands of candidates, yet identifying the right frames seemingly requires understanding the video first. We resolve this circular dependency with a simple observation: cross-modal attention at validation-selected extraction layers in MLLMs already provides query-relevant frame evidence without requiring autoregressive generation. We exploit this property to build DAFS (Dynamic Attention-based Budget-aware Frame Selection), a training-free frame selector. A lightweight MLLM selector, even with only 2B parameters, can extract frame-level evidence by converting selected-layer attention into relevance scores through query-conditioned aggregation. This enables cross-frame comparison without autoregressive decoding. To handle the selector's own context constraint, we formulate the joint allocation of candidate pool size and per-frame token budget as a discrete optimization problem solved by dynamic programming. Under a 32-frame budget, our selector improves over uniform sampling by up to 6.4 points on Video-MME and outperforms prior training-based selectors under matched frame budgets, while generalizing across selector and answerer backbones, and across tasks, without retraining.
Chinese Translation
理解长视频需要从数千个候选帧中选择一个紧凑的帧集，而识别正确的帧似乎需要先理解视频。我们通过一个简单的观察解决了这一循环依赖：在多模态大语言模型（MLLMs）中，验证选择的提取层的跨模态注意力已经提供了与查询相关的帧证据，而无需自回归生成。我们利用这一特性构建了DAFS（动态基于注意力的预算感知帧选择），这是一种无训练的帧选择器。即使只有20亿参数的轻量级MLLM选择器，也可以通过将选定层的注意力转换为相关性分数，通过查询条件聚合提取帧级证据。这使得在没有自回归解码的情况下进行跨帧比较成为可能。为了处理选择器自身的上下文约束，我们将候选池大小和每帧标记预算的联合分配形式化为一个离散优化问题，并通过动态规划求解。在32帧预算下，我们的选择器在Video-MME上比均匀采样提高了最多6.4分，并在匹配帧预算下超越了先前基于训练的选择器，同时在选择器和回答者的骨干网络以及任务之间实现了泛化，而无需重新训练。
cs.CV / 31 / 2607.15698
Hierarchical Specialised Ensembles for Classification of Zebrafish Phenotypes Using the Selected Image Recognition Methods
基于选定图像识别方法的斑马鱼表型分类的层次专门集成模型
Abstract
We propose and evaluate three hierarchical ensemble setups for zebrafish phenotype classification from embryo images. In all setups, stage 1 uses a single four-class classifier to assign images to one of the exclusive phenotypes: Normal, Chorion, Dead, or Other. Images classified as Other are then processed in stage 2, where the ensemble design differs across setups: a single multi-label classifier, two specialized multi-label classifiers, or an ensemble of binary classifiers. We compare these setups using three backbone architectures: ResNet18, ViT, and ConvNeXt. Overall, ConvNeXt achieves the best performance across setups, while the specialized hierarchical ensemble in setup 2 provides the best balance in terms of F1-score. The results show that the proposed specialised hierarchical ensembles are effective for zebrafish phenotype recognition, and suggest that ConvNeXt is particularly useful backbone model.
Chinese Translation
我们提出并评估了三种用于从胚胎图像中分类斑马鱼表型的层次集成设置。在所有设置中，第一阶段使用一个四类分类器将图像分配到四种独占表型之一：正常、卵膜、死亡或其他。被分类为其他的图像随后在第二阶段进行处理，此阶段的集成设计在不同设置中有所不同：一个单一的多标签分类器、两个专门的多标签分类器或一个二元分类器的集成。我们使用三种主干架构进行比较：ResNet18、ViT 和 ConvNeXt。总体而言，ConvNeXt 在所有设置中表现最佳，而设置2中的专门层次集成模型在 F1-score 方面提供了最佳平衡。结果表明，所提出的专门层次集成模型在斑马鱼表型识别中是有效的，并且表明 ConvNeXt 特别适合作为主干模型。
cs.CV / 32 / 2607.15699
GoStop: Reinforcement Learning for Adaptive Temporal Aggregation in Event-Based Feature Tracking
GoStop：用于事件驱动特征跟踪的自适应时间聚合的强化学习
Abstract
Feature tracking plays a fundamental role in understanding scene motion and supports various downstream tasks. Event cameras, with their high temporal resolution and asynchronous sensing, enable low-latency and motion-robust perception, making them well-suited for feature tracking under fast and non-linear motion. However, existing event-based feature tracking methods rely on fixed heuristic rules based on hand-tuning for event accumulation. Such strategies fail to adapt to diverse motion dynamics, leading to degraded performance under abrupt motion changes or low-motion scenarios. In this paper, we model event accumulation as a sequential decision-making problem and introduce reinforcement learning (RL) framework to adaptively control the accumulation process for online event-based feature tracking. Our approach trains a RL agent that decides whether to continue accumulating events or to perform tracking inference based on motion cues. The proposed adaptive temporal agent enables dynamic adaptation to varying motion patterns without relying on hand-crafted rules. Furthermore, we introduce a Dynamic Event-based Tracking (DEFT) dataset with dynamic motion distributions to evaluate the robustness of the feature tracking. Extensive experiments demonstrate that integrating our plug-and-play framework to existing feature tracking methods consistently outperforms heuristic-based approaches, improving robustness under dynamic motion while offering a better balance between tracking accuracy and efficiency. Our project codes and datasets are available at https://github.com/kmax2001/GoSTOP
Chinese Translation
特征跟踪在理解场景运动中发挥着基础性作用，并支持各种下游任务。事件相机凭借其高时间分辨率和异步感知，能够实现低延迟和运动鲁棒感知，使其非常适合在快速和非线性运动下进行特征跟踪。然而，现有的基于事件的特征跟踪方法依赖于基于手动调优的固定启发式规则进行事件累积。这些策略无法适应多样化的运动动态，导致在突发运动变化或低运动场景下性能下降。在本文中，我们将事件累积建模为一个序列决策问题，并引入强化学习（Reinforcement Learning, RL）框架，以自适应地控制在线事件驱动特征跟踪的累积过程。我们的方法训练一个RL代理，根据运动线索决定是继续累积事件还是进行跟踪推断。所提出的自适应时间代理能够在不依赖手工规则的情况下，动态适应不同的运动模式。此外，我们引入了一个具有动态运动分布的动态事件驱动跟踪（Dynamic Event-based Tracking, DEFT）数据集，以评估特征跟踪的鲁棒性。大量实验表明，将我们的即插即用框架集成到现有特征跟踪方法中，始终优于基于启发式的方法，在动态运动下提高了鲁棒性，同时在跟踪准确性和效率之间提供了更好的平衡。我们的项目代码和数据集可在 https://github.com/kmax2001/GoSTOP 获取。
cs.CV / 33 / 2607.15711
Efficient Difficulty-Aware Dynamic Routing for Diffusion-Based Real-World Image Super-Resolution
基于扩散的真实世界图像超分辨率的高效难度感知动态路由
Abstract
Diffusion-based methods have achieved impressive performance in real-world image super-resolution (Real-ISR) by leveraging large pre-trained stable diffusion (SD) models as powerful generative priors. However, these methods still face two key limitations. First, existing SD-based one-step and multi-step Real-ISR approaches adopt a unified processing paradigm for all input samples, ignoring the varying restoration difficulty across images. Second, the aggressive resolution reduction of the VAE in SD models (e.g., 8x downsampling) leads to irreversible loss of fine-scale details, which cannot be recovered by the subsequent diffusion process. To address these limitations, we propose a Difficulty-aware Dynamic Routing (DDR) strategy that overcomes the rigid, one-size-fits-all processing paradigm. Specifically, we first design a difficulty estimator to predict the restoration cost of each input image, enabling automatic assignment to a network of appropriate capacity. Then, we construct a set of Real-ISR networks with varying model capacities by modulating the spatial downsampling ratio of the VAE in the SD backbone, thereby preserving more high-frequency information for challenging cases while maintaining efficiency for simpler inputs. Extensive experiments have demonstrated the superior efficiency and effectiveness of the proposed model compared to recent state-of-the-art methods.
Chinese Translation
基于扩散的方法通过利用大型预训练的稳定扩散（Stable Diffusion, SD）模型作为强大的生成先验，在真实世界图像超分辨率（Real-ISR）中取得了令人印象深刻的性能。然而，这些方法仍面临两个关键限制。首先，现有的基于SD的一步和多步Real-ISR方法对所有输入样本采用统一的处理范式，忽视了图像间恢复难度的差异。其次，SD模型中变分自编码器（Variational Autoencoder, VAE）的大幅分辨率降低（例如，8倍下采样）导致细节的不可逆损失，这些细节无法通过后续的扩散过程恢复。为了解决这些限制，我们提出了一种难度感知动态路由（Difficulty-aware Dynamic Routing, DDR）策略，克服了僵化的“一刀切”处理范式。具体而言，我们首先设计了一个难度估计器，以预测每个输入图像的恢复成本，从而实现自动分配到适当容量的网络。然后，我们通过调节SD主干中VAE的空间下采样比率，构建了一组具有不同模型容量的Real-ISR网络，从而在处理复杂案例时保留更多高频信息，同时对简单输入保持高效性。大量实验表明，与最近的最先进方法相比，所提模型在效率和有效性上具有显著优势。
cs.CV / 34 / 2607.15717
Per-Stroke Temporal Control for Text-to-Motion via Action Units and Action-Detection Guidance
基于动作单元和动作检测引导的文本到运动的逐笔时间控制
Abstract
Text-to-motion models are competent at the action a prompt names but unreliable at when each stroke lands: four punches alternating left and right rarely return four separable strokes. We introduce typed temporal events called Action Units (AUs) that make the individual stroke -- its body track, action class, time window, and impact timing -- an explicit conditioning signal. We ground a frozen text-to-motion backbone on the AU set through a lightweight gated adapter injecting two streams (per-stroke tokens and a per-frame phase channel), and at inference close residual timing errors with a training-free classifier gradient from a frozen frame-level detector. We measure per-stroke control on StrokeBench, whose prompts specify count, ordering, track, and core-frame placement, paired with an audited stroke corpus. AU grounding markedly raises the rate of correctly placed single strokes over the strongest prior interface, at the best motion quality among text-, interval-, and frame-level baselines. The prompted core frame emerges as a further steerable axis.
Chinese Translation
文本到运动模型在执行提示所命名的动作方面表现出色，但在每个笔画落下的时机上却不可靠：四个交替的左拳和右拳很少能返回四个可分离的笔画。我们引入了一种称为动作单元（Action Units, AUs）的类型化时间事件，使得每个笔画——其轨迹、动作类别、时间窗口和冲击时机——成为一个明确的条件信号。我们通过一个轻量级的门控适配器，将冻结的文本到运动骨干网络与动作单元集结合，注入两个流（逐笔令牌和逐帧相位通道），并在推理时通过来自冻结帧级检测器的无训练分类器梯度来缩小残余时间误差。我们在StrokeBench上测量逐笔控制，其提示指定了数量、顺序、轨迹和核心帧位置，并配有经过审核的笔画语料库。动作单元的基础显著提高了正确放置单个笔画的比率，相较于最强的先前接口，在文本、区间和帧级基准中达到了最佳的运动质量。提示的核心帧成为一个进一步可引导的轴。
cs.CV / 35 / 2607.15727
Event3R: Asynchronous-to-Global 3D Reconstruction from Event Camera via Spatial-Temporal Feature Aggregation
Event3R：通过时空特征聚合实现从事件相机到全局的异步三维重建
Abstract
Robust 3D reconstruction is essential for robotics and embodied perception. Recent feed-forward approaches such as DUSt3R have demonstrated impressive progress in dense 3D reconstruction from RGB images, achieving global geometric consistency and strong generalization. However, extending such dense 3D reconstruction to event cameras remains challenging due to their asynchronous, sparse, and highly dynamic nature, as well as the lack of large-scale, well-labeled datasets. In this work, we introduce Event3R, a feed-forward framework that directly maps asynchronous event streams to globally consistent 3D point clouds. Event3R represents incoming events as spatial-temporal voxels, enabling time-aware feature integration through a temporal attention module that enhances the module's temporal feature learning. To further strengthen temporal representation learning and reduce reliance on labeled data, we propose a Masked Bin Modeling (MBM) strategy for self-supervised pre-training, enabling robust temporal representation learning with minimal labeled data, and retain it as an auxiliary fine-tuning objective. In addition, contrastive alignment and consistency regularization losses are incorporated during fine-tuning to reinforce structural correspondence and temporal coherence across views. Extensive experiments on both synthetic and real-world benchmarks demonstrate that Event3R achieves robust, temporally consistent, and globally aligned 3D reconstructions, significantly outperforming existing event-based methods.
Chinese Translation
稳健的三维重建对机器人技术和具身感知至关重要。近期的前馈方法如DUSt3R在从RGB图像进行密集三维重建方面取得了显著进展，实现了全局几何一致性和强大的泛化能力。然而，由于事件相机的异步、稀疏和高度动态的特性，以及缺乏大规模、标注良好的数据集，将这种密集三维重建扩展到事件相机仍然具有挑战性。在本研究中，我们提出了Event3R，这是一种前馈框架，能够直接将异步事件流映射到全局一致的三维点云。Event3R将输入事件表示为时空体素，通过一个时间注意力模块实现时间感知特征的集成，从而增强模块的时间特征学习。为了进一步加强时间表示学习并减少对标注数据的依赖，我们提出了一种掩码箱建模（Masked Bin Modeling，MBM）策略用于自监督预训练，使得在最少标注数据的情况下实现稳健的时间表示学习，并将其作为辅助微调目标。此外，在微调过程中引入对比对齐和一致性正则化损失，以增强视图间的结构对应性和时间一致性。在合成和真实世界基准上的大量实验表明，Event3R实现了稳健、时间一致和全局对齐的三维重建，显著优于现有的基于事件的方法。
cs.CV / 36 / 2607.15732
IoUPD: IoU-Aware Privileged Distillation for Visual Grounding with Multimodal Large Language Models
IoUPD：一种基于IoU的特权蒸馏方法，用于与多模态大型语言模型进行视觉定位
Abstract
Visual grounding with multimodal large language models is commonly formulated as autoregressive coordinate generation, where a model outputs bounding-box coordinates as text given an image and a referring-expression prompt. While this interface is simple and compatible with instruction following, it introduces a mismatch between training and evaluation: training optimizes token-level likelihood over coordinate strings, whereas grounding quality is measured by geometric overlap. We propose IoUPD, an IoU-aware privileged distillation method for coordinate-generating multimodal large language models. IoUPD uses ground-truth boxes not only as coordinate targets, but also as privileged training-time guidance. During training, the student receives the original image and prompt, while a frozen teacher receives a box-marked image and an augmented prompt that indicates the marked region. The student is trained with a supervised fine-tuning anchor and a privileged distillation loss whose token weights reflect both geometric importance and teacher reliability. At inference time, IoUPD requires no box overlay, privileged hint, teacher branch, or additional prediction module. Experiments on standard referring-expression grounding benchmarks show consistent region-level improvements over strong coordinate-generating baselines, demonstrating that ground-truth boxes can provide useful privileged guidance beyond serving as coordinate labels.
Chinese Translation
利用多模态大型语言模型进行视觉定位通常被表述为自回归坐标生成，其中模型根据图像和指代表达提示输出边界框坐标作为文本。尽管这种接口简单且与指令跟随兼容，但它在训练和评估之间引入了不匹配：训练优化坐标字符串的标记级似然，而定位质量则通过几何重叠来衡量。我们提出了IoUPD，一种针对坐标生成的多模态大型语言模型的基于IoU的特权蒸馏方法。IoUPD不仅将真实框作为坐标目标，还作为训练时的特权指导。在训练过程中，学生接收原始图像和提示，而冻结的教师接收标记框的图像和指示标记区域的增强提示。学生通过带有监督微调锚点和特权蒸馏损失进行训练，该损失的标记权重反映了几何重要性和教师可靠性。在推理时，IoUPD不需要框叠加、特权提示、教师分支或额外的预测模块。在标准指代表达定位基准上的实验显示，与强大的坐标生成基线相比，在区域级别上有一致的改善，证明真实框可以提供有用的特权指导，超越作为坐标标签的作用。
cs.CV / 37 / 2607.15734
The Third Competition on Document Forgery Detection on ID-Cards and Passports
第三届身份证和护照文件伪造检测国际竞赛
Abstract
This paper presents a comprehensive analysis of the results from the Third International Competition on Document Forgery Detection on ID-Cards and Passports, which was held across two distinct tracks. Track 1 evaluates a synthetic-data-based ID-PAD system under controlled but diverse conditions, where the winning team, \textit{Incode}, achieves an $AV_{Rank}$ of 27.82%, confirming consistent performance across metrics and highlighting the importance of a balanced, generalizable design. In Track 2, the challenge intensifies with heterogeneous attack scenarios across different domains, where \textit{Incode} again achieved the top position with an $AV_{Rank}$ of 68.71% across thresholds, outperforming some baselines and established methods. These results demonstrate that PAD effectiveness requires not only high accuracy but also consistency across diverse attack types and imaging conditions. The success of this initiative across both tracks underscores the value of collaboration between companies and academic teams. This year, more than \textit{63 teams} were registered, and more than \textit{100 submission models} were evaluated. This competition has evolved into a leading benchmark state-of-the-art in PAD on ID documents, setting the standard for performance, reproducibility, and real-world applicability in secure identity verification.
Chinese Translation
本文对第三届国际身份证和护照文件伪造检测竞赛的结果进行了全面分析，该竞赛分为两个不同的赛道。赛道1在受控但多样的条件下评估基于合成数据的ID-PAD系统，获胜团队 extit{Incode}的$AV_{Rank}$为27.82%，确认了其在各项指标上的一致性表现，并强调了平衡和可推广设计的重要性。在赛道2中，挑战加剧，不同领域的异构攻击场景使得竞争更加激烈， extit{Incode}再次以68.71%的$AV_{Rank}$在各个阈值下取得了第一的位置，超越了一些基准和已有方法。这些结果表明，PAD的有效性不仅需要高准确率，还需要在多样的攻击类型和成像条件下保持一致性。此次活动在两个赛道上的成功凸显了企业与学术团队之间合作的价值。今年，注册的团队超过 extit{63个}，评估的提交模型超过 extit{100个}。该竞赛已发展成为身份证件PAD领域的领先基准，设定了在安全身份验证中性能、可重复性和实际应用的标准。
cs.CV / 38 / 2607.15740
Debiasing Text-to-Image Evaluation via Implicit Cultural Alignment Reward Modeling
通过隐式文化对齐奖励建模消除文本到图像评估的偏见
Abstract
As Text-to-Image (T2I) systems rapidly advance, evaluating the cultural authenticity of synthesized content has become increasingly important for fair and trustworthy generative AI. Existing T2I evaluation metrics and multimodal judges often rely on visual-semantic representations that underrepresent implicit cultural norms, leading to biased preference judgments and the omission of fine-grained cultural cues. In addition, visual question answering (VQA)-based evaluators typically depend on autoregressive text generation, which limits their scalability for real-time reward modeling. To address these limitations, we introduce an Implicit Cultural Alignment Reward Model built upon a lightweight 4.2-billion-parameter Multimodal Large Language Model (MLLM). Our framework integrates an Implicit Cultural Probe with a Skip-connection Cross-Attention (SkipCA) mechanism, enabling late-stage semantic features to directly attend to early-stage visual representations and better preserve culturally salient details. Evaluations on 3,323 challenging and carefully curated image pairs from the CulturalFrames benchmark show that our approach achieves 80.54% pairwise accuracy, with Pearson and Kendall correlation coefficients of 0.546 and 0.377, respectively, outperforming representative vision-language metrics and MLLM-based evaluators. Moreover, by bypassing autoregressive text generation, our model processes each evaluation in 0.21 seconds under our local inference setup, achieving a $10\times$ speedup over standard VQA-based evaluators. These results suggest that the proposed reward model can provide an efficient and culturally aware scalar signal for preference optimization pipelines such as Reinforcement Learning from Human Feedback and Direct Preference Optimization.
Chinese Translation
随着文本到图像（T2I）系统的快速发展，评估合成内容的文化真实性变得越来越重要，以确保生成式人工智能的公平性和可信度。现有的T2I评估指标和多模态评估者通常依赖于视觉-语义表示，这些表示未能充分体现隐式文化规范，导致偏见的偏好判断和细致文化线索的遗漏。此外，基于视觉问答（VQA）的评估者通常依赖自回归文本生成，这限制了它们在实时奖励建模中的可扩展性。为了解决这些限制，我们引入了一种基于轻量级42亿参数多模态大语言模型（MLLM）的隐式文化对齐奖励模型。我们的框架集成了隐式文化探测器和跳跃连接交叉注意力（SkipCA）机制，使得后期的语义特征能够直接关注早期的视觉表示，从而更好地保留文化显著细节。在来自CulturalFrames基准的3,323对具有挑战性且经过精心策划的图像对上的评估结果显示，我们的方法实现了80.54%的成对准确率，Pearson和Kendall相关系数分别为0.546和0.377，超越了代表性的视觉-语言指标和基于MLLM的评估者。此外，通过绕过自回归文本生成，我们的模型在本地推理设置下以0.21秒处理每个评估，相比标准的基于VQA的评估者实现了10倍的加速。这些结果表明，所提出的奖励模型能够为偏好优化管道（如基于人类反馈的强化学习和直接偏好优化）提供高效且具有文化意识的标量信号。
cs.CV / 39 / 2607.15752
Personalized Image Aesthetic Assessment via Preference-rich Sample Mining and Cohort Merging
通过偏好丰富样本挖掘和群体合并实现个性化图像美学评估
Abstract
Personalized Image Aesthetic Assessment (PIAA) aims to predict aesthetic ratings of images that vary across individuals. The aesthetic preferences manifest to different extents across distinct visual stimuli and exhibit cohort-specific patterns. Motivated by the above fact, this paper presents a Multimodal Large Language Model (MLLM)-based approach, which models individual aesthetic preferences by Preference-Rich sample mining and Aesthetically-resonant Cohort merging (PRAC). Specifically, PRAC first identifies preference-rich samples by analyzing both Collective Controversy and Personalized Deviation of images, maximizing the utility of limited user data. Based upon the preference-rich samples, cross-user preference similarities are measured by comparing preference embeddings. Then, a cohort-based model merging strategy, is proposed by aggregating preference patterns from aesthetically-resonant users, which further enhances the personalization for the target individual. Extensive experiments and comparisons on four benchmark PIAA databases demonstrate the superiority of the proposed PRAC model over the state-of-the-arts. The code and model will be public at https://github.com/yzc-ippl/PRAC.
Chinese Translation
个性化图像美学评估（Personalized Image Aesthetic Assessment, PIAA）旨在预测在不同个体中变化的图像美学评分。美学偏好在不同的视觉刺激中表现出不同程度，并展现出特定群体的模式。基于上述事实，本文提出了一种基于多模态大语言模型（Multimodal Large Language Model, MLLM）的方法，通过偏好丰富样本挖掘和美学共鸣群体合并（Preference-Rich sample mining and Aesthetically-resonant Cohort merging, PRAC）来建模个体的美学偏好。具体而言，PRAC首先通过分析图像的集体争议性和个性化偏差来识别偏好丰富样本，从而最大化有限用户数据的效用。在偏好丰富样本的基础上，通过比较偏好嵌入来测量跨用户的偏好相似性。然后，提出了一种基于群体的模型合并策略，通过聚合来自美学共鸣用户的偏好模式，进一步增强目标个体的个性化。对四个基准PIAA数据库进行的大量实验和比较表明，所提出的PRAC模型优于现有的最先进技术。代码和模型将公开在 https://github.com/yzc-ippl/PRAC。
cs.CV / 40 / 2607.15768
GeoChrono: Benchmarking and Rethinking Long-Term Temporal Understanding in Remote Sensing
GeoChrono：遥感中长期时间理解的基准测试与重新思考
Abstract
Remote sensing offers an unparalleled vantage point for observing the Earth's long-term surface evolution, yet it demands that a model not only perceive land cover at isolated moments, but also track changes, memorize evolution histories, and reason across time and space. However, existing studies lack a systematic evaluation that dissects these distinct competencies. To fill this gap, we introduce ChronoBench, a multidimensional benchmark that decomposes this task into four progressive cognitive levels (i.e., Land Cover Perception, Temporal Recognition, Long-Term Memory, and Spatio-Temporal Reasoning). The ChronoBench comprises 12 sub-tasks and 17,689 rigorously validated QA (Question-Answer) pairs. Extensive evaluations reveal that mainstream MLLMs fall drastically behind human experts, with Long-Term Memory emerging as the most critical bottleneck. Motivated by this finding, we further propose GeoChrono, an MLLM with enhanced capabilities for tracing, memorizing, and reasoning about long-term geographic evolution. Leveraging the physical prior that geographic parcels remain spatially fixed while their semantics evolve, we design a Temporal Trajectory Encoder~(TempEnc) that constructs per-location temporal trajectories for dedicated land cover evolution modeling, and we introduce a Coarse-to-Fine Token Compressor~(C2FComp) that adaptively preserves dynamic regions while compressing the static background. To support training, we also construct ChronoInstruct, a 104K-sample instruction-tuning dataset spanning all competency levels for training. GeoChrono achieves state-of-the-art performance on ChronoBench, surpassing the leading commercial MLLMs by over 20%, while C2FComp reduces visual tokens by over 56% while retaining GeoChrono's 94.6% performance. The code and data will be available at https://github.com/IntelliSensing/GeoChrono
Chinese Translation
遥感为观察地球长期表面演变提供了无与伦比的视角，但这要求模型不仅能够在孤立时刻感知土地覆盖，还需跟踪变化、记忆演变历史，并在时间和空间上进行推理。然而，现有研究缺乏对这些不同能力的系统评估。为填补这一空白，我们引入了ChronoBench，一个多维基准，将这一任务分解为四个逐步认知水平（即土地覆盖感知、时间识别、长期记忆和时空推理）。ChronoBench包括12个子任务和17,689个经过严格验证的问答对（QA）。广泛的评估表明，主流的多模态大语言模型（MLLMs）远远落后于人类专家，其中长期记忆被认为是最关键的瓶颈。基于这一发现，我们进一步提出了GeoChrono，一个具备增强能力的MLLM，能够追踪、记忆和推理长期地理演变。利用地理单元在空间上保持固定而其语义不断演变的物理先验，我们设计了一个时间轨迹编码器（Temporal Trajectory Encoder，TempEnc），为专门的土地覆盖演变建模构建每个位置的时间轨迹，并引入了一个粗到细的标记压缩器（Coarse-to-Fine Token Compressor，C2FComp），在压缩静态背景的同时自适应地保留动态区域。为了支持训练，我们还构建了ChronoInstruct，一个包含104K样本的指令调优数据集，涵盖所有能力水平的训练。GeoChrono在ChronoBench上实现了最先进的性能，超过领先的商业MLLMs超过20%，而C2FComp在保持GeoChrono 94.6%性能的同时，将视觉标记减少了超过56%。代码和数据将可在https://github.com/IntelliSensing/GeoChrono获取。
cs.CV / 41 / 2607.15772
SlotMem: Character-Addressable Internal Memory for Narrative Long Video Generation
SlotMem：用于叙事长视频生成的字符可寻址内部记忆
Abstract
Maintaining recurring character identities across scene transitions and long temporal gaps is a central challenge in narrative long video generation. Methods targeting global consistency often retrieve memory using cues that are not aligned with character identity preservation, while recent character-centric variants still rely on coarse frame-level kv memory that entangles identity with incidental visual factors and lacks a continuous update mechanism under limited memory capacity. To address these limitations, we propose \textbf{SlotMem}, a character-addressable internal memory framework for multi-character narrative long video generation. Specifically, SlotMem uses a Character-Semantic Probe to localize character-relevant visual tokens from cross-attention responses, and a Memory Encoder to compress DiT tokens into compact role-wise slot memory. As generation proceeds, a Memory Writer conservatively updates each character's memory with new observations, while Character-Wise Cross-Attention retrieves the role memory and injects it only into localized tokens of the same character. Experiments on multiple narrative long video generation benchmarks show that SlotMem improves long-range character consistency over existing baselines, while maintaining comparable video quality. Our code is available at https://github.com/YilaiLiu-HKU/SlotMem.
Chinese Translation
在场景转换和长时间间隔中保持重复的角色身份是叙事长视频生成中的一个核心挑战。针对全局一致性的方法通常使用与角色身份保持不对齐的线索来检索记忆，而最近的以角色为中心的变体仍然依赖于粗糙的帧级 kv 记忆，这使得身份与偶然的视觉因素纠缠在一起，并且在有限的记忆容量下缺乏连续的更新机制。为了解决这些局限性，我们提出了 extbf{SlotMem}，一个用于多角色叙事长视频生成的字符可寻址内部记忆框架。具体而言，SlotMem 使用字符语义探针从交叉注意力响应中定位与角色相关的视觉标记，并使用记忆编码器将 DiT 标记压缩为紧凑的角色槽记忆。随着生成的进行，记忆写入器保守地用新的观察更新每个角色的记忆，而角色级交叉注意力仅从同一角色的局部标记中检索角色记忆并注入。对多个叙事长视频生成基准的实验表明，SlotMem 在保持可比视频质量的同时，提高了长距离角色一致性。我们的代码可在 https://github.com/YilaiLiu-HKU/SlotMem 获取。
cs.CV / 42 / 2607.15778
Modularized Dynamic-Granularity Video LLM for Multi-Event Long Video Understanding
模块化动态粒度视频大语言模型用于多事件长视频理解
Abstract
Video Large Language Models (Video LLMs) have made significant advancements in various video understanding tasks. However, long-video scenarios remain challenging due to the tension between limited visual token budgets and the need to capture multiple key events. Existing approaches typically process long videos in two stages, i.e., i) select keyframes and ii) perform detailed perception, which exhibit limitations: they lack a modular mechanism for adaptive capacity allocation and self-correction, resulting in unreliable modeling. To tackle these challenges, we propose MoD-VLLM, a novel Modularized Dynamic-Granularity Video LLM framework for multi-event long video understanding, which unifies temporal grounding and semantic understanding iteratively and self-reflectively. Specifically, we propose a Positive-Negative Video Segments Grounding module and a Modularized Dynamic-Granularity Reflection module, which form a closed loop to progressively localize the question-related video segments. The grounding module instructs a Video LLM to distinguish relevant from irrelevant video segments based on the video question. The reflection module employs a modularized scheduler that dynamically selects fine-grained encoding for relevant positive segments to capture detailed perception and coarse-grained encoding for negative segments to maintain global context. We further propose a dynamic-granularity reinforcement learning strategy, allowing MoD-VLLM to learn optimal grounding policies and dynamic granularity visual representation jointly. Moreover, we propose MEventBench, a challenging Multi-Event Long Video Benchmark for complex long video reasoning. Extensive experiments on several long video understanding benchmarks and our MEventBench demonstrate that MoD-VLLM significantly outperforms state-of-the-art baselines.
Chinese Translation
视频大语言模型（Video LLMs）在各种视频理解任务中取得了显著进展。然而，长视频场景仍然面临挑战，因为有限的视觉标记预算与捕捉多个关键事件的需求之间存在矛盾。现有方法通常将长视频处理分为两个阶段，即：i）选择关键帧和 ii）进行详细感知，这些方法存在局限性：缺乏自适应能力分配和自我修正的模块化机制，导致建模不可靠。为了解决这些挑战，我们提出了MoD-VLLM，一种新颖的模块化动态粒度视频大语言模型框架，用于多事件长视频理解，该框架迭代且自我反思地统一了时间定位和语义理解。具体而言，我们提出了一个正负视频片段定位模块和一个模块化动态粒度反思模块，这两个模块形成一个闭环，逐步定位与问题相关的视频片段。定位模块指导视频大语言模型根据视频问题区分相关和无关的视频片段。反思模块采用模块化调度器，动态选择相关正片段的细粒度编码以捕捉详细感知，并对负片段进行粗粒度编码以保持全局上下文。我们进一步提出了一种动态粒度强化学习策略，使MoD-VLLM能够共同学习最佳定位策略和动态粒度视觉表示。此外，我们还提出了MEventBench，这是一个具有挑战性的多事件长视频基准，用于复杂的长视频推理。在多个长视频理解基准和我们的MEventBench上的广泛实验表明，MoD-VLLM显著优于最先进的基线。
cs.CV / 43 / 2607.15779
Multimodal Ambivalence and Hesitancy Recognition via Cross-Attention and Gated Fusion
通过交叉注意力和门控融合进行多模态矛盾与犹豫识别
Abstract
We present a multimodal framework for Ambivalence/Hesitancy (A/H) recognition in video, developed for the ABAW11 challenge at ECCV 2026. The proposed approach fuses textual, acoustic, and visual modalities extracted from the BAH dataset using three pretrained encoders: F2LLM-v2-0.6B for transcripts (1024-d), WavLM-Large for audio (1024-d), and VideoMAE V2 for facial video (768-d). We first establish comprehensive unimodal baselines using classical classifiers (MLP, Random Forest, GBDT), each optimized via Optuna, and obtain a best unimodal Macro F1 of \textbf{0.6659} on the test set using text features alone -- substantially outperforming the zero-shot Video-LLaVA baseline (Macro F1: 0.2827). Building on these baselines, we propose a multimodal fusion architecture that combines bidirectional cross-attention across all three modalities with a Gated Multimodal Unit (GMU), with both architectural and optimization hyperparameters selected through a 50-trial Optuna search. This model achieves a Macro F1 of \textbf{0.7394} on the validation set, a relative improvement of 11.0\% over the best unimodal baseline, confirming that explicit cross-modal interaction captures complementary cues that no single modality provides in isolation. Final predictions on the official, unlabeled private test set are generated using this model and submitted according to the challenge protocol. Code is publicly available at https://github.com/yassineouzar/IUSD_AH/
Chinese Translation
我们提出了一种用于视频中矛盾/犹豫（A/H）识别的多模态框架，该框架是为2026年ECCV的ABAW11挑战而开发的。所提出的方法融合了从BAH数据集中提取的文本、音频和视觉模态，使用三个预训练编码器：用于转录的F2LLM-v2-0.6B（1024维）、用于音频的WavLM-Large（1024维）和用于面部视频的VideoMAE V2（768维）。我们首先使用经典分类器（MLP、随机森林、GBDT）建立全面的单模态基线，每个基线通过Optuna进行优化，并在测试集上仅使用文本特征获得最佳单模态宏F1值为 extbf{0.6659}，显著优于零样本Video-LLaVA基线（宏F1：0.2827）。在这些基线的基础上，我们提出了一种多模态融合架构，该架构结合了所有三种模态之间的双向交叉注意力和一个门控多模态单元（GMU），其架构和优化超参数通过50次Optuna搜索进行选择。该模型在验证集上实现了宏F1值 extbf{0.7394}，相较于最佳单模态基线提高了11.0 extpercnt;，确认了显式的跨模态交互能够捕捉到单一模态无法单独提供的互补线索。最终的预测结果是在官方未标记的私有测试集上使用该模型生成，并按照挑战协议提交。代码可在https://github.com/yassineouzar/IUSD_AH/公开获取。
cs.CV / 44 / 2607.15794
On the Geometry of Learned Representations in Event-Based Multi-Modal Egomotion Estimation
基于事件的多模态自运动估计中学习表示的几何结构
Abstract
Classical approaches to event-based egomotion estimation, including those adopted by the top-performing teams of the ELOPE challenge, rely on geometric optimization frameworks such as contrast maximization, homography estimation, or dense optical flow combined with analytic motion inversion. This work investigates the geometric structure that emerges inside a multi-modal network for egomotion estimation. Event tensors, inertial measurements, and range signals are fused through a cross-modal attention architecture and trained in a batch setting. We analyze the latent space geometry and attention dynamics, showing that (i) embeddings lie on low-dimensional manifolds aligned with motion variables, (ii) attention weights adapt with angular excitation and visual reliability, and (iii) the fused representation recovers classical observability cues. These results bridge analytical estimation theory and modern data-driven fusion.
Chinese Translation
传统的基于事件的自运动估计方法，包括ELOPE挑战赛中表现优异的团队采用的方法，依赖于几何优化框架，如对比度最大化、单应性估计或与解析运动反演相结合的稠密光流。本研究探讨了在多模态网络中用于自运动估计所出现的几何结构。事件张量、惯性测量和距离信号通过跨模态注意力架构进行融合，并在批处理设置中进行训练。我们分析了潜在空间的几何特征和注意力动态，显示出(i) 嵌入位于与运动变量对齐的低维流形上，(ii) 注意力权重随着角度激励和视觉可靠性而调整，以及(iii) 融合表示恢复了经典的可观测性线索。这些结果将分析估计理论与现代数据驱动融合相结合。
cs.CV / 45 / 2607.15806
HybridSim: A Physics-Learning Hybrid Digital Twin for mmWave Human Sensing
HybridSim：一种用于毫米波人类感知的物理学习混合数字双胞胎
Abstract
High-fidelity simulation of mmWave radar signals for dynamic human motion is valuable for developing radar-based human sensing models; yet collecting accurately labeled measurements for a specific deployment site remains expensive. We present HybridSim, a physics-learning hybrid simulator that synthesizes mmWave radar signals from dynamic human meshes under a fixed indoor room configuration, explicitly decoupling propagation into two components. To parameterize the human subject, we use a tri-plane representation to extract human features and a Graph Convolutional Network to stabilize optimization and mitigate gradient instability. The direct signal path is modeled via an inverse-rendering formulation with a microfacet BRDF to capture primary surface reflections. In parallel, the indirect path is approximated by combining 3D Gaussian Splatting with a virtual-receiver geometry to fit and reproduce site-specific multipath interference patterns, achieving substantially lower computational cost than explicit full ray tracing. Experiments in a fixed-room setting show improved agreement with a physically based reference and consistent gains on downstream radar-based human sensing tasks when using HybridSim for site-specific data augmentation.
Chinese Translation
毫米波雷达信号对动态人类运动的高保真模拟对于开发基于雷达的人类感知模型具有重要价值；然而，为特定部署地点收集准确标记的测量数据仍然成本高昂。我们提出了HybridSim，这是一种物理学习混合模拟器，它在固定的室内房间配置下合成动态人类网格的毫米波雷达信号，明确将传播解耦为两个组件。为了参数化人类对象，我们使用三平面表示法提取人类特征，并采用图卷积网络（Graph Convolutional Network）来稳定优化并减轻梯度不稳定性。直接信号路径通过逆渲染公式与微面反射率分布函数（microfacet BRDF）建模，以捕捉主要表面反射。同时，间接路径通过结合三维高斯点云（3D Gaussian Splatting）与虚拟接收器几何体来近似，以拟合和再现特定地点的多径干扰模式，其计算成本显著低于显式全光线追踪。在固定房间设置中的实验表明，使用HybridSim进行特定地点的数据增强时，与基于物理的参考模型的吻合度有所提高，并在下游基于雷达的人类感知任务中获得了一致的提升。
cs.CV / 46 / 2607.15808
Examining the Associations between Visual and Non-Visual Elements and Cyclists' Route Choices for Various Trip Purposes
探讨视觉与非视觉元素与骑行者不同出行目的路线选择之间的关联
Abstract
Understanding cyclist preferences for the characteristics of the built environment is important in promoting sustainable urban transportation and active mobility. Despite previous studies on cyclists' route choices, the influence of visual and non-visual factors on these choices for different trip purposes remains unclear; thus, this paper fills this gap through a data-driven case study in Montreal, Canada. Non-visual factors include socioeconomic factors and two-dimensional environments, while visual factors involve visual perception during cycling and are computed using street view images. The study consists of two parts: one part analyzes spatiotemporal information to explore the non-visual factors between the start and end points of cycling trips, and the other part investigates the discrepancies in distributions of these factors between the shortest path and the actual one. The findings reveal the spatiotemporal characteristics that influence active riding choices, such as increased greenery and lower levels of motorization. These insights can inform the planning of street networks and the development of infrastructure to improve the use of active transportation.
Chinese Translation
了解骑行者对建成环境特征的偏好对于促进可持续城市交通和积极出行至关重要。尽管之前的研究已经探讨了骑行者的路线选择，但视觉和非视觉因素对不同出行目的的影响仍不明确；因此，本文通过在加拿大蒙特利尔的基于数据的案例研究填补了这一空白。非视觉因素包括社会经济因素和二维环境，而视觉因素涉及骑行过程中的视觉感知，并通过街景图像进行计算。研究分为两部分：一部分分析时空信息，以探讨骑行出发点和终点之间的非视觉因素，另一部分则调查这些因素在最短路径与实际路径之间的分布差异。研究结果揭示了影响积极骑行选择的时空特征，如绿化程度的增加和机动车化水平的降低。这些见解可以为街道网络的规划和基础设施的发展提供参考，以改善积极交通的使用。
cs.CV / 47 / 2607.15849
Test-Time Noise Guided Adaptation for Realistic Autoregressive Video Generation
基于测试时噪声引导的适应性方法用于现实自回归视频生成
Abstract
Autoregressive video diffusion models have enabled the generation of arbitrarily long videos by removing conditioning on future frames, thus greatly improving computational efficiency. Yet, they suffer from error accumulation over time, as the denoised sequence gradually drifts away from the conditioning distribution seen during training. Recent advances attempt to reduce this error by anchoring each generated frame to the learned manifold of real ones. However, even when all generated individual frames lie close to the real manifold, there are trajectories which the model lacks sufficient knowledge to continue without exiting it, thus reaching a terminal point. To prevent the model from being trapped in terminal points, we start from the hypothesis that for well-modeled future trajectories the distribution of the predicted noise should match the one of the forward noising process. To enforce such a prior at test time, we introduce Terminal points Avoidance through Noise Guided Optimization (TANGO), which uses the diffusion model as a critic of its own outputs, by predicting one step forward and requiring an isotropic Gaussian noise prediction. We use the deviation from this expected noise distribution to search for an alternative trajectory that does not lead to a terminal point. Our approach achieves a $3.1\%$ absolute improvement on VBench over state-of-the-art, while reducing Fr\'echet Video Distance by $28.3\%$ on average across $15$s videos. Our code is available on https://mever-team.github.io/tango.
Chinese Translation
自回归视频扩散模型通过消除对未来帧的条件限制，实现了任意长度视频的生成，从而大大提高了计算效率。然而，它们在时间上存在误差累积的问题，因为去噪序列逐渐偏离训练期间所见的条件分布。近期的进展尝试通过将每个生成帧锚定到真实帧的学习流形上来减少这种误差。然而，即使所有生成的单个帧都接近真实流形，模型仍然缺乏足够的知识来继续这些轨迹，导致其退出流形，从而达到终点。为了防止模型陷入终点，我们假设对于建模良好的未来轨迹，预测噪声的分布应与前向噪声过程的分布相匹配。为了在测试时强制执行这种先验，我们引入了通过噪声引导优化避免终点（Terminal points Avoidance through Noise Guided Optimization, TANGO），该方法利用扩散模型作为自身输出的评判者，通过预测一步并要求等方差高斯噪声预测。我们利用与这种预期噪声分布的偏差来寻找不导致终点的替代轨迹。我们的方法在VBench上相较于最先进技术实现了$3.1\%$的绝对提升，同时在$15$s视频上平均减少了$28.3\\%$的Fréchet视频距离。我们的代码可在https://mever-team.github.io/tango获取。
cs.CV / 48 / 2607.15851
Von Mises-Fisher Mixture Model with Dynamic Shrinkage for Realistic Test-Time Transduction
具有动态收缩的冯·米塞斯-费舍尔混合模型用于现实测试时传导
Abstract
A range of methods aim to enhance the performance of vision-language models (VLMs) at test time. Among them, transduction has emerged as a promising paradigm due to its strong compatibility and efficiency. However, realistic evaluations often involve highly imbalanced class distributions, which cause performance degradation or even collapse. In this work, we systematically revisit transduction from the perspective of penalized likelihood estimation (PLE), showing that PLE with a KL-divergence anchor term naturally yields an adaptive shrinkage behavior between prior anchors and empirical estimates. From this viewpoint, the brittleness of transductive methods can be attributed to the absence of anchoring mechanism and static modeling of the shrinkage strength. Therefore, we propose Mixture of Von Mises-Fisher Models with Dynamic Shrinkage (MOON). MOON is built upon a mixture of von Mises-Fisher distributions to model feature representations on the unit hypersphere. To handle imbalance, MOON dynamically adjusts the shrinkage strength using zero-shot priors at both instance and class levels. Thus, it suppresses unreliable assignments and prevents harmful updates from outlier classes, thereby mitigating negative transfer. MOON is model-agnostic, training-free, and requires no task-specific hyperparameter tuning. Extensive experiments further validate the advantage of MOON in both performance and efficiency. Our code is available at https://github.com/walawalagoose/MOON
Chinese Translation
一系列方法旨在提高视觉-语言模型（VLMs）在测试时的性能。其中，传导因其强大的兼容性和效率而成为一种有前景的范式。然而，现实评估通常涉及高度不平衡的类别分布，这会导致性能下降甚至崩溃。在本研究中，我们从惩罚似然估计（PLE）的角度系统性地重新审视了传导，表明带有KL散度锚定项的PLE自然产生了先验锚点与经验估计之间的自适应收缩行为。从这个角度来看，传导方法的脆弱性可以归因于缺乏锚定机制和收缩强度的静态建模。因此，我们提出了具有动态收缩的冯·米塞斯-费舍尔模型混合（MOON）。MOON基于冯·米塞斯-费舍尔分布的混合来建模单位超球面上的特征表示。为了处理不平衡，MOON在实例和类别层面上使用零样本先验动态调整收缩强度。因此，它抑制了不可靠的分配，并防止了来自离群类别的有害更新，从而减轻了负迁移。MOON是模型无关的，无需训练，也不需要特定任务的超参数调优。大量实验进一步验证了MOON在性能和效率方面的优势。我们的代码可在 https://github.com/walawalagoose/MOON 获取。
cs.CV / 49 / 2607.15868
EgoExoMoCap: Distributed Ego-Exo Human Motion Capture
EgoExoMoCap：分布式自我-外部人类动作捕捉
Abstract
Human motion capture from head-mounted devices (HMDs) offers a scalable way to acquire real-world human motion and interaction data, which is crucial for applications in embodied AI and VR/AR. Existing approaches focus on either egocentric body tracking, estimating the motion of the subject wearing the device, or exocentric tracking, capturing the movements of people in the wearer's surroundings. So far, these two paradigms have largely been explored in isolation. In this paper, we propose a novel distributed framework that jointly leverages ego- and exocentric multi-modal signals for human motion estimation from HMDs. Unlike traditional motion capture systems requiring bulky multi-camera setups or obtrusive mocap suits, our approach, EgoExoMoCap, is as simple as two (or more) people, each wearing a pair of smart glasses. The method leverages head (plus potentially wrist) tracking signals for accurate estimation of global motion in the 3D world and combines context-aware image features based on DINOv3 to achieve robustness in the presence of noise and occlusions. Extensive experiments on two in-the-wild datasets show that our approach can robustly reconstruct motion even in challenging scenarios.
Chinese Translation
通过头戴式设备（HMD）进行人类动作捕捉提供了一种可扩展的方式来获取现实世界中的人类动作和交互数据，这对于具身人工智能和虚拟/增强现实应用至关重要。现有的方法主要集中在自我中心的身体追踪，即估计佩戴设备的主体的运动，或外部中心的追踪，即捕捉佩戴者周围人群的运动。到目前为止，这两种范式在很大程度上是孤立探索的。本文提出了一种新颖的分布式框架，联合利用自我中心和外部中心的多模态信号来从HMD中进行人类动作估计。与传统的需要庞大多摄像头设置或侵入性动作捕捉服装的动作捕捉系统不同，我们的方法EgoExoMoCap简单到只需两人（或更多），每人佩戴一副智能眼镜。该方法利用头部（以及可能的手腕）追踪信号来准确估计三维世界中的全局运动，并结合基于DINOv3的上下文感知图像特征，以在噪声和遮挡的情况下实现鲁棒性。在两个真实场景数据集上的大量实验表明，我们的方法即使在具有挑战性的场景中也能稳健地重建运动。
cs.CV / 50 / 2607.15887
MDND: Unsupervised Learning Guided by Non-Differentiable Refinement for Shape Correspondence
MDND：一种由非可微精化引导的无监督学习用于形状对应
Abstract
Deep functional map frameworks (DFM) for shape correspondence are powerful, yet fundamentally limited by their reliance on end-to-end differentiability. This constraint prevents the integration of highly accurate, non-differentiable refinement techniques, capping their overall performance, especially on challenging non-isometric shapes. To overcome this, we introduce MDND, a novel DFM paradigm built on the principle of merging differentiable and non-differentiable components. Our framework facilitates unsupervised learning guided by an internal, non-differentiable refinement. Specifically, MDND employs a dual-branch architecture: a non-differentiable refinement branch leverages a novel, multiscale iterative solver to produce highly robust correspondences, acting as a refined target. Concurrently, a fully differentiable branch learns to predict correspondences from features. The entire system is trained end-to-end without supervision by enforcing a consistency loss that compels the differentiable branch to learn from the superior, refined results of the non-differentiable branch. Extensive experiments show that MDND sets a new state-of-the-art, demonstrating remarkable robustness on shapes with non-isometric deformations and topological noise.
Chinese Translation
深度功能映射框架（DFM）在形状对应方面具有强大的能力，但其根本上受到端到端可微性依赖的限制。这一约束阻止了高度准确的非可微精化技术的整合，限制了其整体性能，尤其是在具有挑战性的非等距形状上。为了解决这一问题，我们提出了MDND，一种基于合并可微和非可微组件原则的新型DFM范式。我们的框架促进了由内部非可微精化引导的无监督学习。具体而言，MDND采用双分支架构：一个非可微精化分支利用一种新颖的多尺度迭代求解器生成高度稳健的对应关系，作为精化目标。同时，一个完全可微的分支学习从特征中预测对应关系。整个系统在没有监督的情况下进行端到端训练，通过强制一致性损失，使可微分支从非可微分支的优越精化结果中学习。大量实验表明，MDND设定了新的最先进水平，在具有非等距变形和拓扑噪声的形状上表现出显著的鲁棒性。
cs.CV / 51 / 2607.15889
Hardware-triggered Time Synchronization of Roadside Multi-lidar, Multi-camera Measurement System for Accurate Data Alignment
基于硬件触发的路边多激光雷达、多摄像头测量系统时间同步方法以实现精确数据对齐
Abstract
Accurate temporal alignment of heterogeneous sensors is necessary for reliable environment perception in roadside multi-lidar, multi-camera systems, particularly in dense urban traffic. For this purpose, an open-source, simple, modular, and configurable hardware-triggered time-synchronization circuit is presented in this work to perform temporal alignment or accurate time synchronization between a lidar and multiple cameras. In the designed circuit, a lidar synchronization pulse is used as a reference input, and independently programmable, time-delayed trigger pulses are generated for each camera, allowing flexible adaptation to varying sensor setups and mounting geometries. A series of experiments is conducted on a roadside-mounted perception system comprised of lidar and three cameras, in which the trigger delay is systematically varied, and its impact on spatial-temporal alignment is evaluated. For different classes of road users, the overlap between lidar point cloud measurements and camera measurements is quantified to identify delay configurations that maximize cross-sensor consistency. The proposed circuit is shown to achieve robust and repeatable synchronization while remaining straightforward to deploy, reconfigure, and extend due to its simple and open-source design. Following validation on a three-camera roadside system, the circuit is extended to a vehicle platform with seven cameras and a lidar, providing a low-cost, extensible solution for multi-sensor synchronization across infrastructure and vehicle setups. All hardware circuit design files and source codes are available at https://github.com/shiva-THI/hardware-trigger-time-sync-lidar-cameras.
Chinese Translation
在路边多激光雷达和多摄像头系统中，尤其是在密集城市交通环境中，异构传感器的准确时间对齐对于可靠的环境感知至关重要。为此，本文提出了一种开源、简单、模块化且可配置的硬件触发时间同步电路，以实现激光雷达与多个摄像头之间的时间对齐或准确时间同步。在设计的电路中，激光雷达同步脉冲作为参考输入，为每个摄像头生成独立可编程的时间延迟触发脉冲，从而灵活适应不同的传感器配置和安装几何形状。我们在一个路边安装的感知系统上进行了系列实验，该系统由激光雷达和三台摄像头组成，系统地改变触发延迟，并评估其对时空对齐的影响。对于不同类型的道路使用者，量化激光雷达点云测量与摄像头测量之间的重叠，以识别最大化跨传感器一致性的延迟配置。所提出的电路在实现稳健且可重复的同步的同时，由于其简单且开源的设计，仍然易于部署、重新配置和扩展。在对三摄像头路边系统进行验证后，该电路扩展到一个配备七台摄像头和一台激光雷达的车辆平台，提供了一种低成本、可扩展的多传感器同步解决方案，适用于基础设施和车辆配置。所有硬件电路设计文件和源代码可在 https://github.com/shiva-THI/hardware-trigger-time-sync-lidar-cameras 获取。
cs.CV / 52 / 2607.15890
Exo2EgoPose: Leveraging Exocentric Demonstrations for Vision-Language guided Egocentric 3D Hand Pose Forecasting
Exo2EgoPose：利用外部演示进行视觉-语言引导的自我中心3D手势预测
Abstract
Perceiving multimodal cues and forecasting fine-grained actions from an egocentric (Ego) perspective is vital for applications like robot manipulation. However, previous studies either rely mainly on under-informed visual inputs to predict coarse human motions or follow the VRM/VLA paradigm, which suffers from insufficient robot data and the gap between human and robot embodiments. We observe that 3D hand pose naturally serves as a unified representation to bridge human-robot actions. Hence, we investigate an under-explored Vision-Language guided Egocentric 3D Hand Pose Forecasting (VL-EHPF) task, which aims to predict future Ego 3D hand poses from visual observations, a language instruction, and pose states. To overcome the limited field-of-view and highly dynamic motions in the Ego view, we propose a framework dubbed Exo2EgoPose, which innovatively leverages holistic and stable exocentric (Exo) demonstrations as guidance to compensate for partial and dynamic Ego-view cues. Specifically, we introduce a Dual-level Exocentric Reconstruction Module (DERM), which incorporates the paired Exo videos as supervision to reconstruct their video-level and chunked frame-level representations, thereby modeling spatial contexts and temporal dynamics. Then, the Global-to-Local Modulation Module (GLMM) utilizes the reconstructed hierarchical Exo representations for progressive feature refinement via attention mechanisms and adaptive modulation, enabling comprehensive Exo guidance for accurate Ego hand pose forecasting. Extensive experiments on \textit{AssemblyHands}, \textit{Ego-Exo4D}, and our newly constructed \textit{EgoMe-pose} benchmarks show the superiority of our method, which outperforms state-of-the-art methods by a large margin. Moreover, it demonstrates an effective human-to-robot transfer capability and yields improvements on the \textit{CALVIN} dataset. Code will be released.
Chinese Translation
从自我中心（Ego）视角感知多模态线索并预测细粒度动作对于机器人操作等应用至关重要。然而，以往的研究主要依赖于信息不足的视觉输入来预测粗略的人类动作，或遵循VRM/VLA范式，这在机器人数据不足以及人类与机器人表现之间存在差距的问题上面临挑战。我们观察到，3D手势自然作为一个统一的表示来桥接人类与机器人之间的动作。因此，我们研究了一个尚未深入探索的视觉-语言引导的自我中心3D手势预测（VL-EHPF）任务，旨在从视觉观察、语言指令和姿态状态中预测未来的自我中心3D手势。为了克服自我视角下有限的视野和高度动态的动作，我们提出了一个名为Exo2EgoPose的框架，创新性地利用整体且稳定的外部（Exo）演示作为指导，以补偿部分和动态的自我视角线索。具体而言，我们引入了一个双层外部重建模块（DERM），该模块将配对的Exo视频作为监督，重建其视频级和分块帧级表示，从而建模空间上下文和时间动态。然后，全球到局部调制模块（GLMM）利用重建的层次化Exo表示，通过注意机制和自适应调制进行逐步特征精炼，从而为准确的自我手势预测提供全面的Exo指导。在 extit{AssemblyHands}、 extit{Ego-Exo4D}和我们新构建的 extit{EgoMe-pose}基准上的广泛实验表明我们的方法优越性，显著超越了现有的最先进方法。此外，它展示了有效的人类到机器人转移能力，并在 extit{CALVIN}数据集上取得了改进。代码将会发布。
cs.CV / 53 / 2607.15898
Orbis 2: A Hierarchical World Model for Driving
Orbis 2：一种层次化的驾驶世界模型
Abstract
Current world models operate at a single level of abstraction, with most prioritizing perceptual fidelity while lacking the spatial reasoning and semantic understanding required for real-world downstream tasks. We present a hierarchical driving world model that factorizes future prediction across two levels operating at distinct temporal and abstraction scales: a high-level predictor that forecasts coarse scene structure over extended temporal horizons, and a low-level generator that produces detailed predictions conditioned on the high-level output. This decomposition yields high perceptual fidelity while also capturing strong spatial and semantic representations. We further show that pretraining with a diffusion forcing objective yields substantially richer internal representations than the standard teacher forcing objective, while teacher forcing -- predicting only the next frame from clean context -- produces more stable autoregressive rollouts. We therefore introduce a generic two-stage training paradigm that pretrains the model with diffusion forcing and fine-tunes with teacher forcing, combining the representational benefits of the former with the rollout stability of the latter. Our approach achieves state-of-the-art results across the standard suite of driving world model evaluations on established benchmarks, including long-horizon generation fidelity, steering responsiveness evaluated on counterfactual scenarios, and internal representation quality. Project page with code, demo, checkpoints and qualitative results: https://lmb-freiburg.github.io/orbis2.github.io/
Chinese Translation
当前的世界模型在单一抽象层次上运行，大多数模型优先考虑感知保真度，但缺乏进行现实世界下游任务所需的空间推理和语义理解。我们提出了一种层次化的驾驶世界模型，该模型在两个不同的时间和抽象尺度上对未来预测进行分解：一个高层预测器用于在较长时间范围内预测粗略场景结构，一个低层生成器则基于高层输出生成详细预测。这种分解方法在保持高感知保真度的同时，也捕捉到了强大的空间和语义表示。我们进一步展示了使用扩散强制目标进行预训练所产生的内部表示比标准教师强制目标要丰富得多，而教师强制——仅从干净的上下文中预测下一个帧——则产生了更稳定的自回归展开。因此，我们引入了一种通用的两阶段训练范式，首先使用扩散强制进行模型预训练，然后使用教师强制进行微调，将前者的表示优势与后者的展开稳定性结合起来。我们的方法在已建立基准的标准驾驶世界模型评估中实现了最先进的结果，包括长时间生成保真度、在反事实场景中评估的转向响应性以及内部表示质量。项目页面包含代码、演示、检查点和定性结果： https://lmb-freiburg.github.io/orbis2.github.io/
cs.CV / 54 / 2607.15912
HETA++: Global Structure-from-Motion with Hybrid Explicit Translation Averaging
HETA++：具有混合显式平移平均的全局运动结构重建
Abstract
Global Structure-from-Motion (SfM) offers advantages over incremental methods in terms of efficiency and error distribution. However, the task of translation averaging remains challenging. Many existing methods rely solely on relative translations or feature tracks, which either degrade under collinear camera motion or are susceptible to outliers. In this paper, we propose a novel hybrid explicit translation averaging framework that incorporates both relative translations and feature tracks. Specifically, we first refine the relative translations using global camera rotations and remove globally inconsistent relative translations. Next, we employ convex distance-based objective functions to estimate the initial camera positions and 3D points, followed by refinement using a non-bilinear angle-based objective function. Furthermore, since camera rotations are fixed during translation averaging, inaccurate camera rotations can severely limit the accuracy of camera positions. To address this issue, we then robustly refine both camera rotations and camera positions with selected feature tracks through bounded angle-based refinement and subsequent reprojection-based bundle adjustment. In this step, feature tracks are selected to maintain a balanced spatial distribution and improve optimization efficiency. Finally, we perform a complete bundle adjustment using all reliable feature tracks to refine the camera parameters and 3D points. Extensive experiments on various sequential and unordered real-world datasets demonstrate the superior accuracy, robustness, and scalability of our approach, outperforming state-of-the-art methods in both accuracy and computational efficiency.
Chinese Translation
全局运动结构重建（SfM）在效率和误差分布方面相较于增量方法具有优势。然而，平移平均的任务仍然具有挑战性。许多现有方法仅依赖于相对平移或特征轨迹，这在共线相机运动下会降低效果，或容易受到异常值的影响。本文提出了一种新颖的混合显式平移平均框架，结合了相对平移和特征轨迹。具体而言，我们首先利用全局相机旋转来细化相对平移，并去除全局不一致的相对平移。接下来，我们采用基于凸距离的目标函数来估计初始相机位置和三维点，然后使用基于非双线性角度的目标函数进行细化。此外，由于在平移平均过程中相机旋转是固定的，不准确的相机旋转会严重限制相机位置的准确性。为了解决这个问题，我们通过有界角度的细化和随后的重投影束调整，稳健地细化选定特征轨迹的相机旋转和相机位置。在此步骤中，选择特征轨迹以保持平衡的空间分布并提高优化效率。最后，我们使用所有可靠的特征轨迹进行完整的束调整，以细化相机参数和三维点。在各种顺序和无序的真实世界数据集上进行的广泛实验表明，我们的方法在准确性、鲁棒性和可扩展性方面优于最先进的方法，在准确性和计算效率上均表现出色。
cs.CV / 55 / 2607.15933
Distributional Matching for Vector Quantization: A Unified Theoretical and Empirical Framework
向量量化的分布匹配：统一的理论与实证框架
Abstract
The effectiveness of modern visual representation learning and autoregressive models critically depends on vector quantization (VQ), which discretizes continuous feature representations using a learnable codebook. Despite its widespread use, existing VQ methods often suffer from training instability and codebook collapse, arising from gradient mismatch induced by the straight-through estimator and the under-utilization of code vectors. In this work, we show that both issues can be traced to a fundamental mismatch between the distributions of feature vectors and code vectors, leading to inefficient representation and information loss. Building on this observation, we propose a distributional matching framework for vector quantization. We introduce principled criteria for desirable VQ behavior and demonstrate through theoretical analysis and empirical evaluation that aligning feature and code vector distributions provides a unifying mechanism for mitigating training instability and codebook collapse. We instantiate this framework using a Wasserstein-based objective with an efficient closed-form under a mild Gaussian approximation, and further show that a nonparametric alternative based on maximum mean discrepancy yields comparable performance. Extensive experiments on visual tokenization benchmarks support the effectiveness and robustness of the proposed approach.
Chinese Translation
现代视觉表征学习和自回归模型的有效性在很大程度上依赖于向量量化（VQ），该方法使用可学习的代码本对连续特征表征进行离散化。尽管VQ方法被广泛使用，但现有的VQ方法常常面临训练不稳定和代码本崩溃的问题，这些问题源于直通估计器引起的梯度不匹配以及代码向量的利用不足。在本研究中，我们表明这两个问题可以追溯到特征向量和代码向量分布之间的根本不匹配，导致表征效率低下和信息损失。基于这一观察，我们提出了一种用于向量量化的分布匹配框架。我们引入了理想VQ行为的原则性标准，并通过理论分析和实证评估证明，特征向量和代码向量分布的对齐提供了一种统一机制，以减轻训练不稳定和代码本崩溃的问题。我们使用基于Wasserstein的目标，在温和的高斯近似下，给出了高效的闭式形式，并进一步表明，基于最大均值差异的非参数替代方案能够实现可比的性能。在视觉标记化基准上的大量实验支持了所提方法的有效性和鲁棒性。
cs.CV / 56 / 2607.15936
Handwritten and Printed Text Segmentation via Region-Aware Human-Writing Descriptor Engineering
基于区域感知人类书写描述符工程的手写与印刷文本分割
Abstract
With the increasing demand for reusing paper documents in educational and office settings, accurate segmentation of handwritten and printed text has become a crucial step in document digitization. Although numerous deep learning models have been developed for this task, their high computational cost limits deployment on resource-constrained edge devices. To address this challenge, we present a lightweight framework optimized for efficient performance on devices with severely limited computational capacity. Our approach begins with the Sentence-level Connected Component Segmentation algorithm, aimed at extracting coherent sentence-level segments from document images. We then design a novel Region-aware Handwriting Descriptor (RHD) to capture the intrinsic variability of human handwriting at the sentence level. A simple conventional classifier can then be seamlessly integrated with our designed descriptor, demonstrating strong classification performance for distinguishing handwritten and printed sentence-level text images, highlighting that the proposed descriptor is agnostic to the choice of classifier. Extensive experiments are performed on our self-constructed Multilingual High-Quality Annotated Dataset for Handwritten and Printed Text Segmentation (MAD-HPTS) and a public benchmark PHD-AS, and the experimental results demonstrate that the proposed framework outperforms current state-of-the-art methods in both accuracy and computational efficiency. On MAD-HPTS, our method sacrifices only 1.4% accuracy compared to the leading deep neural network baseline, yet achieves more than 8 times speedup in inference, making it well-suited for lightweight deployment.
Chinese Translation
随着教育和办公环境中对纸质文档重用需求的增加，手写和印刷文本的准确分割已成为文档数字化的重要步骤。尽管为此任务开发了众多深度学习模型，但其高计算成本限制了在资源受限的边缘设备上的部署。为了解决这一挑战，我们提出了一种轻量级框架，优化了在计算能力极为有限的设备上的高效性能。我们的方法首先采用句子级连通组件分割算法，旨在从文档图像中提取连贯的句子级段落。接着，我们设计了一种新颖的区域感知手写描述符（Region-aware Handwriting Descriptor, RHD），以捕捉人类手写在句子级别的内在变异性。然后，可以将一个简单的传统分类器与我们设计的描述符无缝集成，展示出在区分手写和印刷句子级文本图像方面的强大分类性能，突显出所提描述符与分类器的选择无关。我们在自建的多语言高质量手写与印刷文本分割标注数据集（Multilingual High-Quality Annotated Dataset for Handwritten and Printed Text Segmentation, MAD-HPTS）和一个公共基准数据集PHD-AS上进行了广泛实验，实验结果表明，所提框架在准确性和计算效率上均优于当前的最先进方法。在MAD-HPTS上，我们的方法仅牺牲了1.4%的准确性，相比于领先的深度神经网络基线，推理速度提升超过8倍，使其非常适合轻量级部署。
cs.CV / 57 / 2607.15942
More with Less: a Large Scale Remote Sensing VLM with a Simple Recipe
少即是多：一种简单方案的大规模遥感视觉语言模型
Abstract
Remote sensing vision-language models are increasingly expected to support open-ended reasoning over Earth Observation data and a variety of tasks. Most recent progress in this area has been driven by remote-sensing-specific architectural designs, often introducing new encoders, alignment modules, or task-specific fusion mechanisms. In this work, we challenge the necessity of such architectural specialization. We show that a generally capable vision-language model can achieve competitive or state-of-the-art performance at challenging remote sensing benchmarks, provided that it is trained at sufficient scale across diverse data and tasks. Our model uses a single language policy that can either answer directly in text or invoke a localization tool for segmentation and grounding. To train this heterogeneous behaviour, we employ a multi-task reinforcement learning framework with adaptive task rewards covering multiple-choice VQA, free-form VQA, captioning, detection, and segmentation across a large variety of input types. Our approach achieves competitive results across a broad set of benchmarks, including high-resolution, multi-temporal, multi-modal and multi-view tasks. Further, as training data scales, our experiments show consistent improvements across most tasks both in and out of distribution, which correlate with per-task data diversity. These findings suggest that, for remote sensing VLMs, data scale is more important than architectural novelty.
Chinese Translation
遥感视觉语言模型越来越被期望支持对地球观测数据和各种任务的开放式推理。该领域最近的进展主要受到遥感特定架构设计的推动，通常引入新的编码器、对齐模块或特定任务的融合机制。在本研究中，我们质疑这种架构专业化的必要性。我们展示了一种通用的视觉语言模型在具有挑战性的遥感基准测试中能够实现竞争性或最先进的性能，前提是它在多样化的数据和任务上经过充分规模的训练。我们的模型使用单一的语言策略，可以直接以文本形式回答问题或调用定位工具进行分割和定位。为了训练这种异构行为，我们采用了一种多任务强化学习框架，结合适应性任务奖励，涵盖多选视觉问答（VQA）、自由形式视觉问答、图像描述、检测和分割，适用于多种输入类型。我们的方法在广泛的基准测试中取得了竞争性结果，包括高分辨率、多时相、多模态和多视角任务。此外，随着训练数据规模的扩大，我们的实验显示大多数任务在分布内外的一致性改进，这与每个任务的数据多样性相关。这些发现表明，对于遥感视觉语言模型而言，数据规模比架构新颖性更为重要。
cs.CV / 58 / 2607.15995
CanonicalPhys: Pose-Robust Remote Photoplethysmography via Canonical-Space Priors
CanonicalPhys：通过典范空间先验实现姿态鲁棒的远程光电容积描记法
Abstract
Deep remote photoplethysmography (rPPG) attains sub-bpm heart-rate error on frontal, stationary faces yet degrades sharply under head pose: on MMPD, the state-of-the-art FactorizePhys backbone's MAE grows $1.60\times$ from frontal ($|\text{yaw}|{<}15^\circ$) to large-yaw ($|\text{yaw}|{\geq}45^\circ$) frames. We argue that pose is a \emph{coordinate-structural} nuisance rather than a data-augmentation problem: in image coordinates the same pixel maps to different anatomy at different poses, blocking three priors otherwise natural for rPPG, namely the dichromatic reflection model, pulse-phase invariance across skin regions, and the POS/CHROM chromaticity projection, each of which presumes a stable anatomy-to-pixel mapping. We introduce \textbf{CanonicalPhys}, which prepends a differentiable four-point homography that fixes four facial anchors at canonical positions; in this canonical frame the three priors become expressible as a per-pixel Lambertian weight, a cross-ROI temporal consistency loss, and knowledge distillation from windowed POS, none of which adds trainable parameters over the backbone. At an identical parameter count, CanonicalPhys reduces MMPD's frontal-to-large-yaw MAE degradation from $1.60\times$ to $1.33\times$ and flattens the mild-yaw bin from $1.32\times$ to $1.07\times$ (across CanonicalPhys variants), with matched cross-dataset MAE reductions of up to $32\%$ on pose-rich targets. Code: https://github.com/infraface/CanonicalPhys
Chinese Translation
深度远程光电容积描记法（rPPG）在正面静止人脸上达到亚每分钟心率误差，但在头部姿态变化下急剧下降：在MMPD数据集上，最先进的FactorizePhys骨干网络的平均绝对误差（MAE）从正面姿态（$| ext{yaw}|{<}15^ ext{°}$）到大偏航（$| ext{yaw}|{ ext{≥}}45^ ext{°}$）帧增长了$1.60 imes$。我们认为姿态是一个 extit{坐标结构}干扰因素，而不是数据增强问题：在图像坐标中，相同的像素在不同姿态下映射到不同的解剖结构，阻碍了rPPG自然适用的三个先验，即二色反射模型、皮肤区域间脉搏相位不变性以及POS/CHROM色度投影，每个先验都假设稳定的解剖结构与像素的映射。我们引入了 extbf{CanonicalPhys}，它在前面添加了一个可微分的四点单应性变换，将四个面部锚点固定在典范位置；在这个典范框架中，三个先验可以表示为每个像素的朗伯权重、跨区域的时间一致性损失以及来自窗口化POS的知识蒸馏，这些都没有增加骨干网络的可训练参数。在相同的参数数量下，CanonicalPhys将MMPD的正面到大偏航MAE降级从$1.60 imes$降低到$1.33 imes$，并将轻微偏航区间的MAE从$1.32 imes$降低到$1.07 imes$（在CanonicalPhys变体中），在姿态丰富的目标上实现了高达$32 ext{ ext{%}}$的跨数据集MAE减少。代码链接：https://github.com/infraface/CanonicalPhys
cs.CV / 59 / 2607.16007
Beyond Unfolding: 60x Faster One-Stage Unmixing for Closely-Spaced Infrared Small Targets
超越展开：60倍更快的一阶段红外小目标混合分离
Abstract
Due to the optical diffraction limit and long imaging distances, Closely-Spaced Infrared Small Targets (CSIST) typically exhibit energy overlap, manifesting as indistinguishable blobs in infrared images. This ambiguity invalidates the one-to-one mapping assumption of traditional detection, thereby necessitating a paradigm shift towards CSIST Unmixing, which decomposes these blobs into discrete sub-targets. However, the dominant paradigm deep unfolding networks are shackled by the high latency and structural inflexibility intrinsic to their repetitively iterative architecture. To this end, we propose the Fast One-stage CSIST Unmixing Scheme (FOCUS), a one-stage lightweight paradigm which demonstrates that deep unfolding is not necessary. Motivated by the key observation that image super-resolution (SR) and CSIST Unmixing share an isomorphic degradation model, our insight is that it is possible to achieve a paradigm shift from image SR to CSIST Unmixing via completely transforming the label space, loss functions, and evaluation criteria. Specifically, to avoid entangling geometric recovery with artifact suppression, FOCUS adopts a single pass mapping with an internal coarse-to-fine flow that progressively refines target localization from coarse spatial distributions to finer sub-pixel precision. While sparsity regularization suppresses background clutter, it also attenuates target intensities. To compensate for this attenuation of valid signals, flux conservation is introduced as a competing constraint that restores signal energy back to target centers. To the best of our knowledge, this work is the first attempt to address this task via a lightweight one-stage framework without the DUN paradigm. Experiments demonstrate that our method matches or surpasses the state-of-the-art unfolding approaches in both localization and unmixing accuracy, while boosting the inference speed by 60x.
Chinese Translation
由于光学衍射极限和长成像距离，紧密间隔的红外小目标（CSIST）通常表现出能量重叠，在红外图像中呈现为难以区分的斑块。这种模糊性使得传统检测的一对一映射假设失效，因此需要向CSIST混合分离的范式转变，该方法将这些斑块分解为离散的子目标。然而，主导范式深度展开网络受限于其重复迭代架构固有的高延迟和结构不灵活性。为此，我们提出了快速一阶段CSIST混合分离方案（FOCUS），这是一种轻量级的一阶段范式，证明了深度展开并非必要。我们的动机源于一个关键观察，即图像超分辨率（SR）和CSIST混合分离共享同构降解模型，我们的见解是，通过完全转变标签空间、损失函数和评估标准，可以实现从图像SR到CSIST混合分离的范式转变。具体而言，为了避免将几何恢复与伪影抑制纠缠在一起，FOCUS采用单次映射，内部采用粗到细的流动，逐步将目标定位从粗略的空间分布精细化到亚像素精度。虽然稀疏正则化抑制了背景杂波，但它也减弱了目标强度。为补偿有效信号的这种衰减，引入了通量守恒作为竞争约束，以恢复信号能量回到目标中心。根据我们所知，这项工作是首次尝试通过轻量级一阶段框架解决这一任务，而不依赖于DUN范式。实验表明，我们的方法在定位和混合分离精度上与最先进的展开方法相匹配或超越，同时将推理速度提升了60倍。
cs.CV / 60 / 2607.16012
DPNeXt: A Lightweight Multi-Scale Feature Fusion Framework for Efficient ViT-Based Multi-Task Dense Prediction
DPNeXt：一种轻量级多尺度特征融合框架，用于高效的基于ViT的多任务密集预测
Abstract
Multi-Task Learning (MTL) in robotics perception systems supports comprehensive 3D spatial scene understanding by integrating semantic segmentation and depth estimation. While Vision Foundation Models (VFMs) are increasingly adopted as robust feature encoders, existing decoding strategies present a critical bottleneck. To address this, we propose DPNeXt, a streamlined multi-scale feature fusion decoder and efficient alternative to the standard Dense Prediction Transformer (DPT). DPNeXt uses dual depthwise separable inverted bottlenecks to improve frozen VFM utilization through fusion-centric decoding and independent task modularization. To further mitigate negative inductive transfer between tasks, we introduce the Multi-Task Boundary Guidance (MTBG) strategy. Unlike prior boundary-aware methods that add fusion modules or gating, MTBG applies symmetric boundary-focused supervision to encourage geometric consistency without extra annotation or inference cost. Experiments on Cityscapes show that DPNeXt-S outperforms prior state-of-the-art (SOTA) MTL models, while DPNeXt-B further improves the overall performance and achieves the best results among the compared methods. On NYUv2, DPNeXt-B also achieves the best semantic segmentation and depth estimation results among the compared methods while requiring substantially fewer trainable parameters than prior large-scale MTL models. Compared with the standard DPT, DPNeXt-S reduces trainable parameters by 78.6% and achieves the fastest inference speed among the compared models on resource-constrained laptop hardware. The source code, model checkpoints, and a demo video will be made available at https://github.com/kangjehun/DPNeXt.
Chinese Translation
机器人感知系统中的多任务学习（MTL）通过整合语义分割和深度估计，支持全面的三维空间场景理解。尽管视觉基础模型（VFM）越来越多地被采用作为强大的特征编码器，但现有的解码策略却成为了一个关键瓶颈。为了解决这个问题，我们提出了DPNeXt，一种精简的多尺度特征融合解码器，是标准密集预测变换器（DPT）的高效替代方案。DPNeXt使用双深度可分离反向瓶颈，通过以融合为中心的解码和独立任务模块化，提升了冻结VFM的利用率。为了进一步减轻任务之间的负面归纳迁移，我们引入了多任务边界引导（MTBG）策略。与以往添加融合模块或门控的边界感知方法不同，MTBG应用对称的边界聚焦监督，以鼓励几何一致性，而无需额外的标注或推理成本。在Cityscapes上的实验表明，DPNeXt-S超越了之前的最先进（SOTA）MTL模型，而DPNeXt-B进一步提升了整体性能，并在比较方法中取得了最佳结果。在NYUv2上，DPNeXt-B也在比较方法中实现了最佳的语义分割和深度估计结果，同时所需的可训练参数显著少于之前的大规模MTL模型。与标准DPT相比，DPNeXt-S将可训练参数减少了78.6%，并在资源受限的笔记本硬件上实现了最快的推理速度。源代码、模型检查点和演示视频将发布在https://github.com/kangjehun/DPNeXt。
cs.CV / 61 / 2607.16015
PIXIE: A Zero-Shot texture-invariant 6D pose estimation framework for unseen objects with assembly defects
PIXIE：一种针对具有组装缺陷的未见物体的零-shot纹理不变6D姿态估计框架
Abstract
6D pose estimation remains a key challenge in robotics and computer vision, particularly in industrial environments. The deployment of currently available data-driven methods is often limited by resource-intensive data pipelines, reliance on textured 3D models, and sensitivity to geometric deviations caused by damages or assembly defects. We present PIXIE, a zero-shot framework that estimates the 6D pose of an object from an RGB image using only an untextured 3D model. Synthetic depth and normal maps are rendered from sampled reference viewpoints and matched to the query image via a pretrained cross-modality feature matcher. Matched keypoints are back-projected to obtain 2D--3D correspondences for PnP-based pose estimation. Relying exclusively on geometry makes the method inherently robust to lighting and texture variation, while correspondence filtering handles geometric deviations between the model and physical object. We evaluate on widely-used public benchmarks, reporting state-of-the-art results on texture-less objects without object-specific training, and introduce a novel dataset with assembly defects, texture variations, and occlusion to demonstrate real-world applicability.
Chinese Translation
6D姿态估计仍然是机器人技术和计算机视觉中的一个关键挑战，特别是在工业环境中。目前可用的数据驱动方法的部署通常受到资源密集型数据管道、对纹理化3D模型的依赖以及对因损坏或组装缺陷引起的几何偏差的敏感性等因素的限制。我们提出了PIXIE，这是一种零-shot框架，能够仅使用未纹理化的3D模型从RGB图像中估计物体的6D姿态。合成深度图和法线图从采样的参考视点渲染，并通过预训练的跨模态特征匹配器与查询图像进行匹配。匹配的关键点被反投影以获得基于PnP的姿态估计的2D-3D对应关系。该方法完全依赖几何特征，使其在光照和纹理变化方面具有内在的鲁棒性，而对应关系过滤则处理模型与物理物体之间的几何偏差。我们在广泛使用的公共基准上进行了评估，报告了在无物体特定训练的无纹理物体上的最先进结果，并引入了一个包含组装缺陷、纹理变化和遮挡的新数据集，以展示其在实际应用中的可行性。
cs.CV / 62 / 2607.16056
Multi-Modal Semantic Segmentation of Electrolyzer Components for Sustainable Hydrogen Technologies: A Dual-Branch Deep Learning Approach
可持续氢技术中电解槽组件的多模态语义分割：一种双分支深度学习方法
Abstract
Accurate segmentation of electrolyzer materials is essential for automated disassembly, sustainable recycling, and circular manufacturing in hydrogen technologies. However, this task is challenging due to strong visual similarity between materials, spectral overlap, irregular shapes, and severe class imbalance. To address these challenges, we propose an AI-driven dual-branch framework, Hyperspectral-RGB Electrolyzer Materials Network (HREM-Net), that combines hyperspectral imaging (HSI) and RGB images for electrolyzer material segmentation. We implemented several innovative modules, including Efficient Channel Attention, Coordinate Attention, Mobile Inverted Bottleneck blocks, and Atrous Spatial Pyramid Pooling to capture spectral and spatial features from HSI, and RGB images. With an adaptive gated cross-modal fusion module and composite loss function, HREM-Net achieves a mean class accuracy of 91.66% and a mean Intersection over Union (mIoU) of 0.82 on the Electrolyzers-HSI dataset, outperforming baseline segmentation models. Cross-dataset validation on the PCB-Vision dataset demonstrates strong generalization with 96.91% accuracy and 0.93 mIoU. This work poses its potential as an industrial application to improve electrolyzer efficiency, thereby improving the predictive maintenance of hydrogen production.
Chinese Translation
电解槽材料的精确分割对于氢技术中的自动拆解、可持续回收和循环制造至关重要。然而，由于材料之间的强视觉相似性、光谱重叠、不规则形状以及严重的类别不平衡，这项任务具有挑战性。为了解决这些问题，我们提出了一种基于人工智能的双分支框架——高光谱-RGB电解槽材料网络（Hyperspectral-RGB Electrolyzer Materials Network，HREM-Net），该框架结合了高光谱成像（Hyperspectral Imaging，HSI）和RGB图像进行电解槽材料分割。我们实现了多个创新模块，包括高效通道注意力（Efficient Channel Attention）、坐标注意力（Coordinate Attention）、移动反向瓶颈块（Mobile Inverted Bottleneck blocks）和空洞空间金字塔池化（Atrous Spatial Pyramid Pooling），以从HSI和RGB图像中捕获光谱和空间特征。通过自适应门控跨模态融合模块和复合损失函数，HREM-Net在Electrolyzers-HSI数据集上实现了91.66%的平均类别准确率和0.82的平均交并比（mean Intersection over Union，mIoU），超越了基线分割模型。在PCB-Vision数据集上的跨数据集验证显示出强大的泛化能力，准确率达到96.91%，mIoU为0.93。这项工作展示了其作为工业应用的潜力，以提高电解槽效率，从而改善氢生产的预测性维护。
cs.CV / 63 / 2607.16060
ArtChart: A Benchmark for Faithful Artistic Chart Generation with Integrated Text Rendering
ArtChart：一个集成文本渲染的忠实艺术图表生成基准
Abstract
Artistic charts make data memorable and visually engaging, but generating them faithfully demands simultaneously preserving numerical geometry, rendering exact in-image text, binding labels to correct marks, and maintaining coherent artistic style. Current text-to-image and image editing models frequently fail on these coupled constraints, producing distorted geometries, hallucinated text, misbound labels, or over-stylized marks that undermine readability and mathematical integrity. This paper introduces ArtChart, a framework for artistic chart generation with integrated text rendering, encompassing a task definition, benchmark, and evaluation protocol. This is the first work to simultaneously address mathematically faithful chart synthesis, accurate in-image text rendering, and artistic stylization of chart elements. ArtChart features a chart-specific plug-and-play module conditioned on text-free grayscale chart layouts, ensuring mathematical and logical fidelity. A RL learning strategy with OCR accuracy, layout quality, and aesthetic rewards refines generation, while a multi-expert distillation framework resolves inter-reward conflicts through specialized expert optimization. We construct ArtChart-Bench, a bilingual 2K-prompt benchmark spanning four chart types and diverse label formats. We further design ArtChart-Eval, a six-axis evaluation suite covering mathematical logic, text accuracy, text layout, aesthetics, instruction following, and readability, supporting comparison across T2I, image-editing, controllable-generation, and closed-source API models. Extensive experiments demonstrate that ArtChart consistently outperforms open-source baselines, producing charts that are both visually appealing and math faithful.
Chinese Translation
艺术图表使数据更具记忆性和视觉吸引力，但忠实生成这些图表需要同时保持数值几何形状、精确渲染图中文本、将标签绑定到正确的标记，并保持一致的艺术风格。当前的文本到图像和图像编辑模型常常无法满足这些耦合约束，导致几何形状扭曲、幻觉文本、标签错误绑定或过度风格化的标记，从而削弱可读性和数学完整性。本文介绍了ArtChart，一个集成文本渲染的艺术图表生成框架，包括任务定义、基准测试和评估协议。这是首个同时解决数学忠实图表合成、准确图中文本渲染和图表元素艺术风格化的工作。ArtChart具有一个特定于图表的即插即用模块，基于无文本的灰度图表布局，确保数学和逻辑的忠实性。通过结合OCR准确性、布局质量和美学奖励的强化学习策略来优化生成，而多专家蒸馏框架通过专业化专家优化解决奖励间的冲突。我们构建了ArtChart-Bench，一个涵盖四种图表类型和多样标签格式的双语2K提示基准。我们进一步设计了ArtChart-Eval，一个涵盖数学逻辑、文本准确性、文本布局、美学、指令遵循和可读性的六维评估套件，支持在T2I、图像编辑、可控生成和闭源API模型之间的比较。大量实验表明，ArtChart在视觉吸引力和数学忠实性上始终优于开源基线。
cs.CV / 64 / 2607.16065
Spatial Normalization for Cross-Domain Retinal Layer Segmentation in Optical Coherence Tomography
光学相干断层成像中跨域视网膜层分割的空间归一化
Abstract
Retinal layer segmentation in Optical Coherence Tomography (OCT) is a fundamental step for extracting quantitative biomarkers of retinal structure. Indeed, there is a growing interest in the analysis of OCTs in the context of neurodegenerative diseases. However, segmentation remains challenging due to speckle noise, shadowing artifacts, low contrast between adjacent layers, anatomical variability across subjects, and domain shifts arising from different acquisition protocols and clinical populations. While deep learning methods have achieved remarkable performance, their robustness and generalization across heterogeneous datasets remain limited. In this work, we investigate the role of spatial normalization as a preprocessing strategy to mitigate geometric domain shifts and improve the consistency of retinal layer segmentation. Inspired by standard practices in neuroimaging, we introduce a fovea-centered normalization framework that aligns OCT volumes into a common anatomical reference. We perform a comprehensive evaluation of state-of-the-art deep learning architectures. To provide a comprehensive assessment of segmentation quality, we combine conventional overlap-based metrics at B-scan level with topology-aware metrics at A-scan level and thickness-based measures at the en-face level. In cases where a ground truth is not available, we propose topology violation quantitative metrics that do not require ground truth annotations and a thickness-based qualitative assessment that captures structural consistency and clinically relevant patterns at the en-face level. The results demonstrate the importance of spatial normalization in OCT segmentation pipelines toward the development of robust and clinically meaningful retinal analysis tools, enabling reliable biomarker extraction and downstream computational analysis in neurodegenerative research.
Chinese Translation
光学相干断层成像（OCT）中的视网膜层分割是提取视网膜结构定量生物标志物的基础步骤。实际上，在神经退行性疾病的背景下，对OCT的分析越来越受到关注。然而，由于散斑噪声、阴影伪影、相邻层之间的低对比度、受试者之间的解剖变异性以及因不同采集协议和临床人群引起的领域转变，分割仍然面临挑战。尽管深度学习方法取得了显著的性能，但它们在异构数据集上的鲁棒性和泛化能力仍然有限。在本研究中，我们探讨了空间归一化作为一种预处理策略的作用，以减轻几何领域转变并提高视网膜层分割的一致性。受到神经影像学标准实践的启发，我们引入了一种以中央凹为中心的归一化框架，将OCT体积对齐到一个共同的解剖参考。我们对最先进的深度学习架构进行了全面评估。为了全面评估分割质量，我们结合了基于B扫描的传统重叠度量、基于A扫描的拓扑感知度量和基于面扫描的厚度度量。在缺乏真实标注的情况下，我们提出了不需要真实标注的拓扑违反定量度量和一种基于厚度的定性评估，捕捉在面扫描层面上结构一致性和临床相关模式。结果表明，空间归一化在OCT分割流程中的重要性，为开发鲁棒且具有临床意义的视网膜分析工具奠定了基础，使得在神经退行性研究中能够可靠地提取生物标志物并进行后续的计算分析。
cs.CV / 65 / 2607.16076
HCIG: A Hierarchical Cross-Modal Incongruity Graph Network for Multimodal Sarcasm and Cyberbullying Detection
HCIG：一种用于多模态讽刺和网络欺凌检测的层次交叉模态不一致图网络
Abstract
Multimodal sarcasm and cyberbullying detection remain challenging because the intended meaning often emerges from incongruity between textual and visual information rather than from either modality alone. Existing multimodal approaches primarily rely on feature fusion or cross-modal attention, which may not effectively capture hierarchical semantic inconsistencies across different levels of representation. To address this limitation, this paper proposes HCIG (Hierarchical Cross-modal Incongruity Graph Network), a novel framework that models cross-modal incongruity at token, phrase, and global levels using graph attention networks and adaptively integrates these representations through a learned hierarchical attention mechanism. As a complementary architecture, we also introduce GCCN (Graph-based Cross-modal Contradiction Network), which performs graph-based reasoning using contradiction-aware pooling for efficient multimodal interaction learning. The proposed models are evaluated on the MMSD sarcasm benchmark and the MultiBully cyberbullying dataset, together with comprehensive ablation studies and cross-task transfer experiments. Experimental results demonstrate that HCIG achieves the best performance on MMSD with 85.74% accuracy and 85.29% macro-F1, while GCCN attains the highest macro-F1 (68.66%) on MultiBully and HCIG achieves the highest accuracy (69.62%) and bullying-class F1 (74.90%). The findings demonstrate that hierarchical multi-granularity incongruity modeling provides more effective multimodal reasoning than conventional fusion strategies, offering a robust framework for sarcasm and cyberbullying detection in social media.
Chinese Translation
多模态讽刺和网络欺凌检测仍然面临挑战，因为意图的意义往往源于文本信息与视觉信息之间的不一致，而非单一模态。现有的多模态方法主要依赖特征融合或交叉模态注意力，这可能无法有效捕捉不同层次表示之间的层次语义不一致。为了解决这一局限性，本文提出了HCIG（层次交叉模态不一致图网络），这是一个新颖的框架，利用图注意力网络在标记、短语和全局层面建模交叉模态不一致，并通过学习的层次注意力机制自适应地整合这些表示。作为补充架构，我们还引入了GCCN（基于图的交叉模态矛盾网络），该网络使用矛盾感知池化进行图基推理，以实现高效的多模态交互学习。所提模型在MMSD讽刺基准和MultiBully网络欺凌数据集上进行了评估，并进行了全面的消融研究和跨任务迁移实验。实验结果表明，HCIG在MMSD上取得了最佳性能，准确率为85.74%，宏观F1为85.29%；而GCCN在MultiBully上获得了最高的宏观F1（68.66%），HCIG则在准确率（69.62%）和欺凌类别F1（74.90%）上表现最佳。研究结果表明，层次多粒度不一致建模提供了比传统融合策略更有效的多模态推理，为社交媒体中的讽刺和网络欺凌检测提供了一个稳健的框架。
cs.CV / 66 / 2607.16077
Adaptive Contrast Enhancement and Optimised Feature Matching for RootSIFT-Based Palm-Vein Recognition
基于RootSIFT的掌静脉识别中的自适应对比度增强与优化特征匹配
Abstract
Palm-vein recognition is a highly secure biometric modality due to the uniqueness and subcutaneous nature of vein patterns. However, low contrast in palm-vein images, caused by NIR light scattering and sensor limitations, remains a significant challenge. To address this, we propose the Intensity-Limited Adaptive Contrast Stretching with Bidirectional Gaussian-weighted Overlapping Tiles (ILACS-BGOT) method, an enhancement of the previously developed ILACS with Layered Gaussian-weighted Overlapping Tiles (ILACS-LGOT) technique. ILACS enhances local contrast, while BGOT mitigates blocky artefacts. This study further integrates RootSIFT features with KNN+RT and incorporates the previously introduced Mean and Median Distance (MMD) filter to investigate the parameter variations of both MMD and RT, and their impact on recognition performance. A comprehensive analysis was conducted across three benchmark datasets (CASIA, PolyU, and PUT), using 42 combinations of MMD filter thresholds and RT values. Results were evaluated using EER and Accuracy. Findings reveal that higher template sizes improve performance, while varying MMD thresholds reflect dataset-specific rotational variations. The proposed system demonstrates superior generalisability, achieving significant improvements in both EER and Accuracy over existing methods. Furthermore, the underlying ILACS-BGOT mechanism suggests potential applicability beyond palm vein recognition to other biometric modalities such as finger vein and palmprint recognition, and more generally to low-contrast image enhancement across computer vision applications.
Chinese Translation
掌静脉识别是一种高度安全的生物识别方式，因其静脉图案的独特性和皮下特性而受到重视。然而，由于近红外（NIR）光散射和传感器限制，掌静脉图像的低对比度仍然是一个重大挑战。为了解决这个问题，我们提出了一种强度限制自适应对比度拉伸与双向高斯加权重叠瓦片（ILACS-BGOT）的方法，这是对之前开发的带有分层高斯加权重叠瓦片（ILACS-LGOT）技术的增强。ILACS增强了局部对比度，而BGOT则减轻了块状伪影。本研究进一步将RootSIFT特征与KNN+RT结合，并引入之前提出的均值和中位数距离（MMD）滤波器，以研究MMD和RT的参数变化及其对识别性能的影响。我们在三个基准数据集（CASIA、PolyU和PUT）上进行了全面分析，使用了42种MMD滤波器阈值和RT值的组合。结果通过等错误率（EER）和准确率进行评估。研究发现，较大的模板尺寸提高了性能，而不同的MMD阈值反映了数据集特定的旋转变化。所提出的系统展现出更好的泛化能力，在EER和准确率上相较于现有方法有显著提升。此外，ILACS-BGOT机制的基础表明其潜在应用不仅限于掌静脉识别，还可扩展到其他生物识别方式，如指静脉和掌纹识别，以及更广泛的低对比度图像增强在计算机视觉应用中的应用。
cs.CV / 67 / 2607.16094
How Do VLMs Fail? Vision-Operation Misalignment in Compositional VQA
视觉-操作不一致性：组合视觉问答中的 VLM 失败机制
Abstract
Compositional visual question answering requires Vision-Language Models (VLMs) to execute multiple reasoning operations like object selection, spatial relation resolution, and attribute verification. Despite strong aggregate performance, the mechanistic basis of VLM failures on this task remains underexplored. To address this gap, we analyze vision-operation misalignment in VLMs by examining how failures relate to specific reasoning operations and the internal computational pathways through which they arise and propagate. We introduce an Operation-centric mechanistic framework that decomposes VLM failures by both the reasoning operation where they originate and the internal computational pathway through which they propagate. Our analysis reveals four mechanistically distinct failure modes: grounding failure, reasoning failure, attribute extraction failure, and language prior dominance failure. Each characterized by a unique relationship between visual grounding strength and answer correctness. Through three complementary causal interventions applied across all transformer layers, we further demonstrate a pathway dissociation: grounding failures route exclusively through the feedforward network, reasoning failures route through late-layer attention, and attribute extraction failures localize to the answer-position feedforward computation. This dissociation demonstrates that different failure types require fundamentally different corrective strategies, providing a principled foundation for targeted improvements to VLM reliability in multimedia reasoning.
Chinese Translation
组合视觉问答要求视觉-语言模型（VLM）执行多种推理操作，如对象选择、空间关系解析和属性验证。尽管整体表现强劲，但 VLM 在此任务上的失败机制尚未得到充分探讨。为了解决这一空白，我们通过分析 VLM 中的视觉-操作不一致性，研究失败与特定推理操作之间的关系，以及这些失败是如何产生和传播的内部计算路径。我们提出了一种以操作为中心的机制框架，通过推理操作的来源和传播的内部计算路径来分解 VLM 的失败。我们的分析揭示了四种机制上不同的失败模式：基础失败、推理失败、属性提取失败和语言先验主导失败。每种模式都表现出视觉基础强度与答案正确性之间的独特关系。通过在所有变压器层上应用三种互补的因果干预，我们进一步展示了路径的分离：基础失败仅通过前馈网络传播，推理失败通过后层注意力传播，而属性提取失败则局限于答案位置的前馈计算。这种分离表明，不同类型的失败需要根本不同的纠正策略，为 VLM 在多媒体推理中的可靠性提供了有原则的改进基础。
cs.CV / 68 / 2607.16105
Attention-Guided Saliency Maps for Interpreting Visualization Literacy in VLMs
基于注意力的显著性图用于解释视觉语言模型中的可视化素养
Abstract
Understanding how vision-language models (VLMs) interpret data visualizations remains an open problem, and is increasingly important as these models are used for analytical tasks where reliable reasoning is essential. We introduce a lightweight, diagnostic saliency map method tailored for text generation over images using transformer models, the current state-of-the-art models in visualization interpretation. Our approach aggregates the language model's attention over the visual tokens across all heads and layers, then maps this attention back onto the vision encoder's patch grid to localise it over the image, producing a direct correspondence between each generated answer token and the image regions it attended to. This yields fast, gradient-free saliency maps that expose how VLMs allocate focus across visual elements during answer generation, enabling inspection of whether model attention aligns with semantically relevant components. We evaluate our approach using a deletion metric which validates the causal faithfulness of our saliency maps to the model's behavior.
Chinese Translation
理解视觉语言模型（VLMs）如何解释数据可视化仍然是一个未解决的问题，并且随着这些模型在需要可靠推理的分析任务中的应用，这一问题变得越来越重要。我们提出了一种轻量级的诊断显著性图方法，专为使用变换器模型进行图像文本生成而设计，这些模型是当前可视化解释的最先进模型。我们的方法聚合了语言模型在所有头部和层次上对视觉标记的注意力，然后将这种注意力映射回视觉编码器的补丁网格，以便在图像上进行定位，从而在每个生成的答案标记与其关注的图像区域之间建立直接对应关系。这产生了快速、无梯度的显著性图，揭示了VLMs在答案生成过程中如何分配对视觉元素的关注，使得可以检查模型的注意力是否与语义相关的组件对齐。我们使用删除度量评估我们的方法，该度量验证了我们显著性图对模型行为的因果忠实性。
cs.CV / 69 / 2607.16128
Toward Semantic Communication for Real-time Mobile 3D Reconstruction
面向实时移动3D重建的语义通信
Abstract
Real-time mobile 3D reconstruction is fundamental to many emerging applications such as autonomous navigation and digital twin construction, where a moving platform continuously captures an image stream and transmit to a computing server for scene understanding. Unlike offline reconstruction, camera poses and scene geometry are estimated on-the-fly during acquisition, making multi-view consistency a real-time requirement and rendering geometric estimation highly sensitive to communication-induced distortions. Semantic communication (SemCom) transmits compact semantic information, offering a promising way to preserve task-critical data over unreliable links. However, existing designs are optimized at the image or single-view level and without providing explicit reliability information for geometric estimation, limiting their applicability to real-time mobile 3D reconstruction. In this context, we propose a SemCom framework for real-time mobile 3D reconstruction. The framework includes a semantic transceiver that outputs a reconstructed image alongside a pixel-wise confidence map, quantifying the reliability of each region. We further introduce a confidence-guided geometric estimation method, incorporating confidence into RANSAC-based pose initialization and bundle adjustment to reduce the influence of unreliable regions and enhance robustness under noisy channels. Simulations show that, compared to existing SemCom and traditional seperate source and channel coding, our framework maintains high image quality while significantly improving pose estimation accuracy and 3D structural consistency.
Chinese Translation
实时移动3D重建是许多新兴应用的基础，如自主导航和数字双胞胎构建，其中移动平台持续捕获图像流并传输到计算服务器以进行场景理解。与离线重建不同，摄像机姿态和场景几何是在采集过程中实时估计的，这使得多视图一致性成为实时需求，并使几何估计对通信引起的失真高度敏感。语义通信（Semantic Communication, SemCom）传输紧凑的语义信息，为在不可靠链路上保留任务关键数据提供了有前景的方法。然而，现有设计在图像或单视图层面上进行了优化，并未为几何估计提供明确的可靠性信息，从而限制了其在实时移动3D重建中的适用性。在此背景下，我们提出了一种用于实时移动3D重建的SemCom框架。该框架包括一个语义收发器，该收发器输出重建图像及像素级置信度图，量化每个区域的可靠性。我们进一步引入了一种置信度引导的几何估计方法，将置信度融入基于RANSAC的姿态初始化和束调整中，以减少不可靠区域的影响并增强在噪声信道下的鲁棒性。模拟结果表明，与现有的SemCom和传统的分离源和信道编码相比，我们的框架在保持高图像质量的同时，显著提高了姿态估计的准确性和3D结构的一致性。
cs.CV / 70 / 2607.16154
CLIFE: Camera-LiDAR Fusion Framework for Edge-Deployable Roadside VRU Perception
CLIFE：用于边缘部署的路边脆弱道路用户感知的相机-激光雷达融合框架
Abstract
Reliable roadside perception of vulnerable road users (VRUs) remains challenging under occlusions, variable lighting, and diverse weather conditions, particularly under strict edge-computing and latency constraints. Existing multi-sensor fusion systems rely on cloud or server-grade infrastructure, creating a deployment gap at real-world intersections. We present CLIFE, an edge-native camera-LiDAR fusion framework that integrates targetless online calibration and lightweight late-fusion tracking entirely on a single embedded device, without cloud offloading. CLIFE adaptively refines camera-LiDAR alignment on demand and performs multi-sensor fusion and track association with O(N log N) per-frame cost. We deploy CLIFE across 12 signalized intersections in Chattanooga and conduct an in-depth evaluation at a representative intersection using synchronized camera-LiDAR data that spans diverse daytime, nighttime, and weather conditions. Our experiments demonstrate that the fusion architecture substantially enhances the perceptual range and robustness of the individual sensors under varied environmental and traffic conditions. The late-fusion core operates at 53.2 FPS on the Jetson AGX Thor, ensuring high throughput for real-time intersection-scale applications. By centering perception at the edge, CLIFE provides a deployable foundation for downstream safety applications, while reducing bandwidth and calibration overhead for agencies operating multi-intersection corridors.
Chinese Translation
在遮挡、光照变化和多样天气条件下，可靠的路边脆弱道路用户（VRUs）感知仍然面临挑战，尤其是在严格的边缘计算和延迟限制下。现有的多传感器融合系统依赖于云或服务器级基础设施，导致在实际交叉口的部署存在差距。我们提出了CLIFE，一个边缘原生的相机-激光雷达融合框架，能够在单一嵌入式设备上完全实现无目标在线校准和轻量级后融合跟踪，而无需云端卸载。CLIFE根据需求自适应地优化相机-激光雷达对齐，并以每帧O(N log N)的成本执行多传感器融合和轨迹关联。我们在查塔努加的12个信号交叉口部署了CLIFE，并在一个具有代表性的交叉口进行深入评估，使用同步的相机-激光雷达数据，涵盖了多样的白天、夜晚和天气条件。我们的实验表明，该融合架构显著增强了在不同环境和交通条件下各个传感器的感知范围和鲁棒性。后融合核心在Jetson AGX Thor上以53.2 FPS的速度运行，确保了实时交叉口规模应用的高吞吐量。通过将感知中心放在边缘，CLIFE为下游安全应用提供了可部署的基础，同时减少了多交叉口走廊运营机构的带宽和校准开销。
cs.CV / 71 / 2607.16165
An Exam for Active Observers
主动观察者的考试
Abstract
Human vision is a closed loop: gaze is continuously redirected by intermediate hypotheses rather than a single snapshot. Decades of psychophysics and cognitive science have argued that this active observation is essential for a wide range of tasks. Whether today's multimodal large language models (MLLMs) exercise active observation is an empirical question that current vision-language benchmarks do not answer. We introduce ActiveVision, a benchmark that makes active observation measurable for MLLMs, comprising 17 tasks across 3 categories. Tasks are designed to force repeated visual perception rather than a single static description. Frontier MLLMs collapse on ActiveVision: the highest-scoring model we evaluate, GPT-5.5 at the highest exposed reasoning-effort tier, solves only 10.6% of items and scores zero on 11 of the 17 tasks, and even Claude Fable 5, despite topping most reasoning and coding leaderboards, solves just 3.5%, far behind three human participants who average 96.1%. Furthermore, much of the gap persists even when models write and run their own vision code: such code is unreliable on realistic imagery, and catching its failures itself requires the active perception the models lack. Together, these results indicate that current MLLMs lack robust active visual observation, motivating architectures and training objectives that close the perception-reasoning loop.
Chinese Translation
人类视觉是一个闭环：注视点不断被中间假设重新引导，而不是单一快照。几十年的心理物理学和认知科学表明，这种主动观察对于广泛的任务至关重要。当前的多模态大型语言模型（MLLMs）是否进行主动观察是一个实证问题，而现有的视觉-语言基准并未对此进行解答。我们引入了ActiveVision，这是一个使主动观察可测量的基准，包含17个任务，分为3个类别。任务设计旨在迫使进行重复的视觉感知，而不是单一的静态描述。前沿的MLLMs在ActiveVision上表现不佳：我们评估的得分最高的模型GPT-5.5在最高的推理努力层级中，仅解决了10.6%的项目，并且在17个任务中有11个得分为零，尽管Claude Fable 5在大多数推理和编码排行榜上名列前茅，但它的解决率仅为3.5%，远远落后于三位人类参与者的平均96.1%。此外，即使模型编写并运行自己的视觉代码，差距仍然存在：这些代码在真实图像上不可靠，而捕捉其失败本身需要模型所缺乏的主动感知。综合来看，这些结果表明当前的MLLMs缺乏稳健的主动视觉观察，促使我们设计能够缩小感知-推理闭环的架构和训练目标。
cs.CV / 72 / 2607.16181
Vision-Language Assistant for Emotional Reactions to Risky Driving
情感反应的风险驾驶视觉-语言助手
Abstract
This study introduces a vision-language pipeline that detects risky driving behaviors and generates emotionally expressive responses to support driver awareness and comfort. Although vision-language models have advanced perception and reasoning in autonomous driving, existing systems rarely consider the emotional dimension or real-world user experience. Keep Yelling Assistant (KYA) detects high-risk driving maneuvers in real time, such as sudden cut-ins. It then produces emotional responses through a large language model tailored to driver preferences. The framework comprises two core modules. The vision module uses YOLOv8 variants to detect nearby vehicles and identify risky behaviors such as sudden cut-ins. Key driving metrics, including relative distance, speed, and projected reach time, are extracted and normalized to produce a structured behavior log. The language module processes this log with user-defined emotional tone settings, such as neutral, humorous, and analytical, and generates verbal reactions using state-of-the-art large language models, including ChatGPT-4o, Claude 3, Gemini 2.5, and Copilot. We evaluated the proposed system using dashcam videos containing risky driving behaviors and a user study involving 108 participants. Participants selected preferred response styles, and the large language models were evaluated based on emotional alignment. All models received favorable ratings, although preferences varied across personas. Notably, the combination of YOLOv8s and ChatGPT-4o achieved the highest score of 4.29 out of 5.00. By integrating real-world perception with emotionally adaptive dialogue, KYA introduces a new paradigm for emotionally intelligent in-vehicle artificial intelligence. It offers promising directions for improving safety, trust, and emotional well-being in both conventional and autonomous vehicles.
Chinese Translation
本研究介绍了一种视觉-语言管道，能够检测风险驾驶行为并生成情感表达的反应，以支持驾驶员的意识和舒适度。尽管视觉-语言模型在自动驾驶中的感知和推理方面取得了进展，但现有系统很少考虑情感维度或现实世界用户体验。Keep Yelling Assistant (KYA) 实时检测高风险驾驶操作，如突然插入。随后，它通过一个针对驾驶员偏好的大型语言模型生成情感反应。该框架由两个核心模块组成。视觉模块使用 YOLOv8 变体检测附近车辆并识别风险行为，如突然插入。提取并标准化相对距离、速度和预计到达时间等关键驾驶指标，以生成结构化的行为日志。语言模块根据用户定义的情感语调设置（如中性、幽默和分析）处理该日志，并使用最先进的大型语言模型（包括 ChatGPT-4o、Claude 3、Gemini 2.5 和 Copilot）生成口头反应。我们使用包含风险驾驶行为的行车记录仪视频和涉及 108 名参与者的用户研究评估了所提出的系统。参与者选择了偏好的反应风格，大型语言模型的情感一致性进行了评估。所有模型均获得了良好的评分，尽管不同角色的偏好有所不同。值得注意的是，YOLOv8s 和 ChatGPT-4o 的组合获得了 5.00 满分中的 4.29 的最高分。通过将现实世界的感知与情感自适应对话相结合，KYA 引入了一种情感智能车载人工智能的新范式。它为提高传统和自动驾驶车辆的安全性、信任度和情感福祉提供了有希望的方向。
cs.CV / 73 / 2607.16189
Searching Videos as Trees: Self-Correcting Agents for Grounded Long Video QA
将视频视为树：用于基础长视频问答的自我纠正智能体
Abstract
Grounded long-video question answering (Grounded LVQA) requires answering a question about a long video while localizing the short evidence interval that supports the answer. Recent agentic methods frame this task as multi-turn exploration with a single crop_video(start, end) action, which supports coarse-to-fine narrowing but provides no primitive for fine-to-coarse backtracking. As a result, these agents typically converge prematurely and cannot recover from an early mistake. We propose VideoTreeSearch (VTS), a framework that casts grounded LVQA as iterative self-correcting search over an adaptive temporal tree. VTS constructs a non-uniform tree from visual scene boundaries so that each node corresponds to a semantically coherent segment, and trains an agent to navigate the tree through four discrete operations: zoom_in, zoom_out, shift, and answer. These operations expose backtracking and recovery as explicit, learnable primitives rather than implicit behaviors. To train this navigation, we introduce a trajectory synthesis pipeline that produces multi-step paths through the tree, including deliberate detours into incorrect branches followed by recovery. We use these trajectories for supervised fine-tuning, followed by reinforcement learning with grounding and answer-accuracy rewards. On three Grounded LVQA benchmarks (CG-Bench, Haystack-LVBench, Haystack-Ego4D), VTS outperforms the strongest prior agentic methods by +12.5 mIoU on CG-Bench and +7.4 T-F1 on Haystack-Ego4D. The learned policy also transfers to general long-video QA, surpassing all prior agentic baselines on Video-MME, MLVU, and LVBench by up to +7.1 accuracy points. Ablations confirm that self-correcting hierarchical search is the central mechanism behind these gains: removing either adaptive descent or explicit backtracking substantially degrades performance. Code is available at https://github.com/CeeZh/VTS.
Chinese Translation
基础长视频问答（Grounded LVQA）需要回答关于长视频的问题，同时定位支持答案的短证据区间。近期的智能体方法将此任务框架设定为通过单一的 crop_video(start, end) 动作进行多轮探索，这种方法支持从粗到细的缩小，但未提供从细到粗的回溯原语。因此，这些智能体通常会过早收敛，无法从早期错误中恢复。我们提出了视频树搜索（VideoTreeSearch, VTS），一个将基础 LVQA 视为在自适应时间树上进行迭代自我纠正搜索的框架。VTS 从视觉场景边界构建一个非均匀树，使得每个节点对应一个语义上连贯的片段，并训练一个智能体通过四个离散操作：zoom_in、zoom_out、shift 和 answer 来导航树。这些操作将回溯和恢复显式化为可学习的原语，而不是隐式行为。为了训练这种导航，我们引入了一个轨迹合成管道，生成通过树的多步路径，包括故意偏离错误分支后再进行恢复。我们使用这些轨迹进行监督微调，随后进行强化学习，结合基础和答案准确性的奖励。在三个基础 LVQA 基准（CG-Bench、Haystack-LVBench、Haystack-Ego4D）上，VTS 在 CG-Bench 上比最强的先前智能体方法提高了 +12.5 mIoU，在 Haystack-Ego4D 上提高了 +7.4 T-F1。学习到的策略也转移到一般的长视频问答，在 Video-MME、MLVU 和 LVBench 上超越所有先前的智能体基线，准确率提高了最多 +7.1 个百分点。消融实验确认，自我纠正的层次搜索是这些提升的核心机制：去除自适应下降或显式回溯会显著降低性能。代码可在 https://github.com/CeeZh/VTS 获取。
cs.CV / 74 / 2607.16190
FVAttn: Adaptive Sparse Attention with Runtime Load Balancing for Video Generation
FVAttn：具有运行时负载平衡的视频生成自适应稀疏注意力
Abstract
Video Diffusion Transformers process long spatio-temporal sequences, making self-attention the main bottleneck in high-resolution video generation. Training-free sparse attention reduces this cost, but adaptive Top-$p$ routing creates uneven per-head workloads under multi-GPU sequence parallelism. The resulting workload heterogeneity turns sparse attention into a rank-level straggler problem. We present \method{}, a training-free sparse-attention system that improves the distributed execution efficiency of adaptive sparse attention under multi-GPU sequence parallelism. \method{} uses Top-$p$ routing, a Top-$k$ safety floor, and video-aware block organization as the sparse-routing frontend, then repairs the materialized mask at runtime. Runtime Load Balancing migrates a small number of heavy heads via P2P communication to shorten the current critical path. Slack-Aware Sparse Augmentation fills residual non-critical-rank slack with additional high-value blocks, while overlap hides scheduling and migration overhead behind existing computation. On step-distilled Wan2.2 I2V, \method{} reduces average load imbalance from 1.34 to 1.08 and delivers a $4.41\times$ attention speedup over FlashAttention, while achieving a $2.02$--$2.11\times$ DiT inference speedup with competitive video quality.
Chinese Translation
视频扩散变换器处理长时间空间序列，使得自注意力成为高分辨率视频生成的主要瓶颈。无训练的稀疏注意力降低了这一成本，但自适应 Top-$p$ 路由在多 GPU 序列并行下产生了不均匀的每个头的工作负载。由此产生的工作负载异质性将稀疏注意力转变为等级级别的滞后问题。我们提出了 extit{method}，一种无训练的稀疏注意力系统，旨在提高自适应稀疏注意力在多 GPU 序列并行下的分布式执行效率。 extit{method} 使用 Top-$p$ 路由、Top-$k$ 安全底线和视频感知的块组织作为稀疏路由前端，然后在运行时修复物化掩码。运行时负载平衡通过 P2P 通信迁移少量重头，以缩短当前的关键路径。Slack-Aware Sparse Augmentation 用额外的高价值块填充剩余的非关键等级松弛，而重叠则隐藏了现有计算后的调度和迁移开销。在步骤蒸馏的 Wan2.2 I2V 上， extit{method} 将平均负载不平衡从 1.34 降低到 1.08，并提供了相较于 FlashAttention 的 $4.41 imes$ 注意力加速，同时在竞争的视频质量下实现了 $2.02$--$2.11 imes$ DiT 推理加速。
cs.CV / 75 / 2607.16192
MotionForesight: Re-purposing Video Models for Future 3D Scene-Flow Prediction
运动前瞻：重新利用视频模型进行未来3D场景流预测
Abstract
Humans can infer how objects are likely to move from passive observation: a cup may be lifted, a drawer may slide, and a lid may rotate shut. Such predictions expose the physical consequences of interaction needed to act in the real world. We study how to learn this anticipation from ordinary monocular videos of human-object interaction. Given a short observed video context, MotionForesight predicts future 3D trajectories for points on the manipulated object. This casts interaction prediction as object-centered 3D motion forecasting without any assumptions on the object properties. Our key insight is that video prediction models already encode rich priors about how objects move during human interactions. We redirect these priors from pixel prediction toward future 3D scene flow. We start from a dense 3D tracker built on a pretrained video model, generate pseudo-ground-truth tracks from complete clips, and train the forecaster using only the observed frames. We replace future RGB and geometry with learned mask latents and train a lightweight adapter to turn the retrospective tracking representation into a forward predictor, while freezing the large video and tracking components. Using just 40k human videos and no auxiliary inputs such as language, MotionForesight generalizes across diverse out-of-distribution objects, environments, viewpoints, and interactions. It also outperforms substantially larger models that use over a million training videos. These results show that we can efficiently re-purpose video priors into explicit geometric forecasts for embodied intelligence. https://motionforesight.github.io/
Chinese Translation
人类可以通过被动观察推断物体可能的运动方式：杯子可能被举起，抽屉可能滑动，盖子可能旋转关闭。这种预测揭示了进行现实世界互动所需的物理后果。我们研究如何从普通的单目视频中学习这种预期，视频中展示了人类与物体的互动。在给定的短暂观察视频上下文中，MotionForesight预测被操作物体上点的未来3D轨迹。这将互动预测视为以物体为中心的3D运动预测，而不对物体属性做任何假设。我们的关键见解是，视频预测模型已经编码了关于物体在与人类互动过程中如何运动的丰富先验知识。我们将这些先验知识从像素预测转向未来3D场景流。我们从一个基于预训练视频模型构建的稠密3D跟踪器开始，从完整的视频片段生成伪真实轨迹，并仅使用观察到的帧训练预测器。我们用学习到的掩码潜变量替代未来的RGB和几何信息，并训练一个轻量级适配器，将回顾性跟踪表示转化为前向预测器，同时冻结大型视频和跟踪组件。仅使用4万个人类视频，并且没有语言等辅助输入，MotionForesight能够在多样的分布外物体、环境、视角和互动中进行泛化。它的表现也超越了使用超过一百万个训练视频的更大模型。这些结果表明，我们可以有效地将视频先验知识重新利用为具体现实智能的明确几何预测。
cs.CV / 76 / 2607.16193
Knowing the Self, Understanding the World: A Dual-Cognition Benchmark for UAV Spatio-temporal Reasoning with MLLMs
认识自我，理解世界：基于多模态大语言模型的无人机时空推理双认知基准
Abstract
Multimodal large language models have achieved strong performance across diverse vision-language tasks, yet their capabilities in UAV scenarios remain insufficiently explored. Recent UAV-oriented benchmarks have begun to evaluate MLLMs in aerial scenarios, but they typically focus on scene understanding, event recognition, or navigation completion, rather than jointly assessing the dual-cognition capability required for UAV agents: reasoning about both the UAV's own state and the external environment in multiview spatio-temporal contexts. To address this gap, we present UAV-DualCog, a benchmark for aerial multiview spatio-temporal reasoning built on this dual-cognition perspective. UAV-DualCog includes both image and video tasks to jointly evaluate self-state and environment-state reasoning, while requiring spatial or temporal grounding beyond discrete answer prediction. We also develop an automated pipeline that constructs data from scene-level semantic point clouds, yielding a scalable benchmark with diverse scenes, hundreds of landmarks, and thousands of QA samples. Extensive evaluations show that current MLLMs remain far from reliable in UAV dual cognition. Self-state reasoning, viewpoint transformation, precise spatial grounding, and temporal interval localization are persistent bottlenecks, and additional validation with thinking/frontier models and a human baseline confirms that the benchmark is understandable to humans but challenging for existing models. We further construct UAV-DualCog-Train from disjoint scenes and show through a lightweight optimization probe that it provides useful structured supervision, suggesting its value not only as an evaluation benchmark but also as a data resource for advancing MLLM-based UAV agents. Project website and supplementary materials: https://uav-dualcog.lozumi.com
Chinese Translation
多模态大语言模型在多样的视觉-语言任务中取得了强劲的表现，但其在无人机场景中的能力仍然未得到充分探索。近期针对无人机的基准测试开始评估多模态大语言模型在空中场景中的表现，但通常关注于场景理解、事件识别或导航完成，而非联合评估无人机代理所需的双认知能力：在多视角时空上下文中推理无人机自身状态和外部环境。为填补这一空白，我们提出了UAV-DualCog，这是一个基于双认知视角构建的空中多视角时空推理基准。UAV-DualCog包括图像和视频任务，以共同评估自我状态和环境状态的推理，同时要求在离散答案预测之外进行空间或时间的基础定位。我们还开发了一个自动化管道，从场景级语义点云构建数据，形成一个具有可扩展性的基准，涵盖多样的场景、数百个地标和数千个问答样本。广泛的评估表明，当前的多模态大语言模型在无人机双认知方面仍远未可靠。自我状态推理、视角转换、精确空间定位和时间间隔定位是持续的瓶颈，额外的验证与思维/前沿模型以及人类基线的比较确认了该基准对人类是可理解的，但对现有模型却具有挑战性。我们进一步从不相交的场景构建了UAV-DualCog-Train，并通过轻量化优化探针展示它提供了有用的结构化监督，表明它不仅作为评估基准具有价值，也作为推动基于多模态大语言模型的无人机代理的数据资源。项目网站及补充材料： https://uav-dualcog.lozumi.com
cs.AI / 1 / 2607.15280
GraphDx: A Cost-Aware Knowledge-Enhanced Multi-Agent Framework for Sequential Diagnosis
GraphDx：一种成本意识的知识增强多智能体框架用于序列诊断
Abstract
Sequential diagnosis requires balancing diagnostic accuracy against resource costs through iterative information gathering. Existing Large Language Model (LLM) approaches exhibit a critical knowledge-reasoning gap: despite encoding extensive medical knowledge, they struggle to reason systematically under cost constraints, often resorting to excessive testing. We propose GraphDx, a knowledge-enhanced framework with two core innovations. First, we design an automated pipeline that leverages LLMs to construct Medical Diagnosis Knowledge Graphs (MDKGs) with quantized typicality, action-centric topology, and dual-objective attributes for both diagnostic relevance and cost-sensitivity. Second, we introduce three collaborative agents (Perception, Reasoning, and Decision) where the Perception and Decision Agents handle language understanding and generation, while the Reasoning Agent performs deterministic evidence scoring and cost-aware planning on the MDKG. Experiments on MedQA and MIMIC-IV across three LLM backbones (DeepSeek-V3, Kimi-k2, Llama-3.3) show that GraphDx improves diagnostic success rates from 50--68% to 79--93% while reducing test costs by 20--54%, providing a robust, economical, and interpretable solution for automated clinical diagnosis.
Chinese Translation
序列诊断需要在诊断准确性与资源成本之间进行平衡，通过迭代信息收集来实现。现有的大型语言模型（LLM）方法存在显著的知识推理差距：尽管编码了广泛的医学知识，但在成本约束下，它们难以系统性地进行推理，常常 resorting to excessive testing。我们提出了GraphDx，这是一种知识增强框架，具有两个核心创新。首先，我们设计了一个自动化管道，利用LLMs构建具有量化典型性、以行动为中心的拓扑结构以及双重目标属性（既包括诊断相关性又包括成本敏感性）的医学诊断知识图（MDKGs）。其次，我们引入了三个协作代理（感知、推理和决策），其中感知代理和决策代理处理语言理解和生成，而推理代理在MDKG上执行确定性证据评分和成本意识规划。在MedQA和MIMIC-IV上的实验，涵盖了三个LLM基础模型（DeepSeek-V3、Kimi-k2、Llama-3.3），结果表明GraphDx将诊断成功率从50-68%提高到79-93%，同时将测试成本降低了20-54%，为自动化临床诊断提供了一种稳健、经济且可解释的解决方案。
cs.AI / 2 / 2607.15281
Causal-Audit: Explicit and Auditable Graph-based Reasoning via Target-Aware Causal Chain Construction
因果审计：通过目标感知因果链构建实现显式和可审计的图基推理
Abstract
Causal and intervention-based question answering is fundamental to advancing large language models (LLMs) toward reasoning beyond surface-level correlations and understanding underlying causal mechanisms. However, existing LLM-based methods often rely on implicit language-level reasoning, resulting in opaque causal assumptions, unverifiable reasoning paths, and fragile predictions under complex interventions, particularly in context-free settings. In this paper, we propose an explicit and auditable causal reasoning framework for context-free intervention-based question answering. Our method formulates causal inference as structured reasoning over an explicit causal graph through four modular stages, rather than implicit end-to-end prediction. A key innovation is a target-aware causal graph construction strategy that treats the target variable as a core constraint during graph expansion, effectively suppressing irrelevant variables, spurious causal relations, and reasoning noise. We further introduce a path-level causal evidence aggregation mechanism that combines multiple causal paths while modeling both reinforcing and counteracting effects, enabling robust decision-making beyond single-chain reasoning. Extensive experiments on three benchmarks demonstrate that our framework consistently outperforms existing LLM-based methods while providing interpretable and auditable causal reasoning traces.
Chinese Translation
基于因果和干预的问题回答是推动大型语言模型（LLMs）向超越表面相关性、理解潜在因果机制发展的基础。然而，现有的基于LLM的方法往往依赖于隐式的语言层推理，导致因果假设不透明、推理路径不可验证以及在复杂干预下的脆弱预测，尤其是在无上下文的环境中。本文提出了一种显式且可审计的因果推理框架，用于无上下文的基于干预的问题回答。我们的方法将因果推断表述为在显式因果图上的结构化推理，分为四个模块阶段，而不是隐式的端到端预测。一个关键创新是目标感知的因果图构建策略，该策略在图扩展过程中将目标变量视为核心约束，有效抑制无关变量、虚假因果关系和推理噪声。我们进一步引入了一种路径级因果证据聚合机制，该机制在建模强化和抵消效应的同时，结合多条因果路径，使得决策过程超越单链推理，变得更加稳健。在三个基准测试上的大量实验表明，我们的框架在提供可解释和可审计的因果推理轨迹的同时，始终优于现有的基于LLM的方法。
cs.AI / 3 / 2607.15314
Cura 1T: Specialized Model for Agentic Healthcare
Cura 1T：专门针对代理医疗的模型
Abstract
Healthcare spans high-stakes communication, expert reasoning, and workflow execution, yet specialized LLMs that cover these use cases together remain limited. A healthcare model must handle patient consultation, clinical reasoning over text and images, interactive diagnosis, and electronic health record (EHR) tool use. These capabilities fail in different ways, and a narrow update for one task can degrade another. We present Cura 1T, a healthcare-specialized LLM trained through a human-gated self-evolution loop. In each evolution round, a training agent plans a target capability, trains the model, evaluates benchmark trajectories, and refines the data mixture from observed failures. This data-centered loop improves the model through targeted synthetic and curated examples rather than a single generic medical-data update. Across the healthcare evaluation suite, Cura 1T ranks at or near the top among frontier baselines, while remaining competitive on out-of-domain reasoning and agentic benchmarks.
Chinese Translation
医疗保健涉及高风险沟通、专家推理和工作流程执行，但涵盖这些用例的专门大型语言模型（LLMs）仍然有限。医疗模型必须处理患者咨询、文本和图像的临床推理、互动诊断以及电子健康记录（EHR）工具的使用。这些能力在不同方面存在缺陷，针对某一任务的狭窄更新可能会降低其他任务的性能。我们提出了Cura 1T，这是一种通过人类引导的自我进化循环训练的医疗专用大型语言模型。在每个进化轮次中，训练代理规划目标能力，训练模型，评估基准轨迹，并根据观察到的失败优化数据混合。这个以数据为中心的循环通过针对性的合成和策划示例来改进模型，而不是依赖单一的通用医疗数据更新。在医疗评估套件中，Cura 1T在前沿基准中排名靠前，同时在领域外推理和代理基准上保持竞争力。
cs.AI / 4 / 2607.15367
AnovaX: A Local, Multi-Agent Voice Assistant with LLM Planning, Typed Executors, and Adaptive Recovery
AnovaX：一个具有LLM规划、类型化执行器和自适应恢复的本地多智能体语音助手
Abstract
Desktop voice assistants are still dominated by cloud pipelines that ship raw audio off the machine and expose a fixed set of skills. We describe AnovaX, a small local-first assistant that runs entirely on the user's computer and treats the desktop itself as its action surface. A single Python process wires together a wake-word gate, a speech pipeline, an LLM planner (Gemini) that emits a JSON plan of tool calls, a whitelist-and-denylist safety layer, a multi-agent orchestrator that translates each plan into typed child agents on a bounded thread pool, and an adaptive recovery loop that takes over whenever a core step fails. Every tool corresponds to a specialized agent class (AppAgent, TypingAgent, BrowserAgent and six others) with its own timeout, retry policy, and shared-resource locks. A recursive MetaAgent lets the planner delegate a sub-goal back to itself, capped at two levels of nesting. The recovery loop uses a compact ReAct-style prompt and hides Gemini's latency behind speculative execution of read-only tools. A companion Flask server exposes a phone-friendly remote over the local WiFi, mirrors every agent lifecycle event to the phone in real time, and streams the laptop's screen back over MJPEG so the user can watch remote commands land as they run. The point of the project is less to compete with Siri or Alexa than to show that a legible, few-thousand-line assistant is enough to open apps, type into them, run searches, coordinate concurrent actions, recover from single-step failures, and be driven entirely from a phone in another room -- without the LLM ever touching the keyboard.
Chinese Translation
桌面语音助手仍然主要依赖于将原始音频发送到云端的管道，并提供固定的一组技能。我们描述了AnovaX，一个完全在用户计算机上运行的小型本地优先助手，将桌面本身视为其操作界面。一个单一的Python进程将唤醒词门控、语音处理管道、发出工具调用JSON计划的LLM规划器（Gemini）、白名单和黑名单安全层、多智能体协调器（将每个计划转换为在有限线程池上的类型化子代理），以及一个自适应恢复循环（在核心步骤失败时接管）连接在一起。每个工具对应一个专门的代理类（AppAgent、TypingAgent、BrowserAgent及其他六个），每个类都有自己的超时、重试策略和共享资源锁。递归的MetaAgent允许规划器将子目标委托回自身，嵌套层级最多为两层。恢复循环使用紧凑的ReAct风格提示，并通过对只读工具的推测执行来隐藏Gemini的延迟。一个配套的Flask服务器通过本地WiFi暴露出适合手机的远程接口，实时将每个代理的生命周期事件镜像到手机上，并通过MJPEG流回笔记本电脑的屏幕，以便用户可以在命令执行时观看远程命令的到达。该项目的重点不是与Siri或Alexa竞争，而是展示一个可读的、几千行的助手足以打开应用程序、输入内容、进行搜索、协调并发操作、从单步失败中恢复，并完全通过另一个房间的手机进行驱动——而不需要LLM接触键盘。
cs.AI / 5 / 2607.15388
Precise but Uncoupled: Reviewer Precision Does Not Guarantee Critique Uptake in Multi-Agent Math Reasoning
精准但不耦合：审稿人精准度并不保证多智能体数学推理中的批评采纳
Abstract
Many math- and science-oriented agent systems use hierarchical designs with specialized reviewer roles, assuming that a dedicated review stage should help turn wrong candidates into correct ones. We test this assumption on 4,181 verifier-grounded Omni-MATH problems using matched gpt-oss-120b actors. Collaboration adds little on the easiest tiers, but from tier 4 onward the gains open sharply; in this harder regime, broadcast-style peer discussion reaches higher final accuracy than a planner-executor-reviewer pipeline (PER). We ask whether this gap is explained by reviewer quality or by whether critique changes the next answer the protocol carries forward. It is not explained by reviewer precision alone: PER's reviewer is more precise than broadcast's (0.861 vs. 0.644), yet evaluator-verified useful critique is much less likely to change the next candidate and produces lower reviewer-guided repair. These results show that reviewer detection quality and critique uptake are empirically separable. Within matched PER interventions, forcing explicit acknowledgment lowers final accuracy, while embedding reviewer guidance directly in the solver's working context partially improves follow-through without closing the gap. Overall, reviewer-centric evaluation can overstate system quality: a protocol may spot errors well yet still fail to solve more problems if it does not act on those critiques.
Chinese Translation
许多面向数学和科学的智能体系统采用层次化设计，设有专门的审稿人角色，假设专门的审查阶段应有助于将错误的候选者转变为正确的候选者。我们在4181个基于验证者的Omni-MATH问题上测试了这一假设，使用匹配的gpt-oss-120b智能体。协作在最简单的层次上几乎没有增加效果，但从第4层开始，收益急剧增加；在这一更困难的层次中，广播式的同行讨论达到了比计划者-执行者-审稿人管道（PER）更高的最终准确性。我们探讨这一差距是否由审稿人质量解释，或是批评是否改变了协议所传递的下一个答案。仅凭审稿人精准度无法解释这一差距：PER的审稿人比广播的更精准（0.861对0.644），然而经过评估者验证的有用批评却更不可能改变下一个候选者，并产生较低的审稿人引导修复。这些结果表明，审稿人检测质量和批评采纳在经验上是可分离的。在匹配的PER干预中，强制明确承认会降低最终准确性，而将审稿人指导直接嵌入解题者的工作上下文中则部分改善了后续执行，但未能缩小差距。总体而言，审稿人中心的评估可能夸大系统质量：一个协议可能能够很好地发现错误，但如果不对这些批评采取行动，仍然可能无法解决更多问题。
cs.AI / 6 / 2607.15418
DrawingVQA: A Real-World Benchmark for Multi-Depth Visual-Textual Reasoning on Construction Drawings
DrawingVQA：针对建筑图纸的多深度视觉-文本推理的现实世界基准
Abstract
We introduce DrawingVQA, the first benchmark designed to evaluate multimodal large language models (MLLMs) on real-world construction drawings -- a core media in architecture, civil, and many other engineering practices. Unlike natural images or schematic floor plans, construction drawings fuse abstract geometry, symbolic notation, tabular data, annotations, and domain-specific text, forming a uniquely complex visual-textual domain core to engineering workflows. DrawingVQA bridges this gap with 33 "Issued for Construction" drawings and 92 expertly curated question-answer pairs, spanning three reasoning depths: perceptual understanding, contextual interpretation, and domain-expert reasoning. To evaluate model capabilities, we present a dual categorization framework to jointly analyze performance across seven construction-engineering and four MLLM capability dimensions -- the first to explicitly map engineering workflows to AI reasoning competencies. Evaluations of state-of-the-art MLLMs reveal a substantial gap between model and expert performance, particularly at higher reasoning depths. This benchmark lays a foundation for domain-specialized multimodal reasoning to allow for advancement on integration of AI-driven understanding and real-world engineering workflows.
Chinese Translation
我们介绍了DrawingVQA，这是第一个旨在评估多模态大型语言模型（MLLMs）在现实世界建筑图纸上的基准——建筑、土木工程以及许多其他工程实践中的核心媒介。与自然图像或示意性平面图不同，建筑图纸融合了抽象几何、符号标记、表格数据、注释和特定领域文本，形成了一个在工程工作流程中独特复杂的视觉-文本领域。DrawingVQA通过33幅“用于施工”的图纸和92对专家精心策划的问题-答案对，填补了这一空白，涵盖了三种推理深度：感知理解、上下文解释和领域专家推理。为了评估模型能力，我们提出了一个双重分类框架，以共同分析在七个建筑工程和四个MLLM能力维度上的性能——这是首次明确将工程工作流程映射到人工智能推理能力。对最先进的MLLMs的评估显示，模型与专家性能之间存在显著差距，尤其是在更高的推理深度上。该基准为领域专业化的多模态推理奠定了基础，以促进人工智能驱动的理解与现实世界工程工作流程的整合。
cs.AI / 7 / 2607.15439
Do Coding Agents Need Executable World Models, Simplification, and Verification to Solve ARC-AGI-3?
编码代理是否需要可执行的世界模型、简化和验证以解决ARC-AGI-3？
Abstract
Our previous ARC-AGI-3 agent bundled executable world modeling, scheduled simplification, and exact replay verification, leaving unclear which idea accounted for its performance. We address this attribution question with four nested Codex-based agents: a textual baseline; a flexible-interface executable world model without replay verification; the same executable model with scheduled simplification; and a fixed-interface verification treatment that retains simplification and requires exact reproduction of recorded observations. The main study evaluates all four agents with gpt-5.4 and gpt-5.5 at high and xhigh reasoning effort on the public ARC-AGI-3 games. Exploratory follow-ups evaluate the textual and verification variants with gpt-5.6-sol at xhigh and max. The most robust result is that every agent variant improves with a stronger model and with greater reasoning effort. Within each model-effort setting, differences among variants are smaller than anticipated, while the effects of individual components vary across settings. Requiring a persistent executable deliverable is not universally beneficial: the textual variant outperforms the flexible-interface executable variant in both gpt-5.5 settings. Simplification improves performance in three of the four model-effort settings, with the weakest setting as the only exception. The complete verification treatment ranks first in all four settings, although it uses substantially more resources. In the gpt-5.6-sol follow-up, the verification variant fully solves every public game at both reasoning efforts, achieves about 99% RHAE, and uses fewer than half the total actions of the human baseline. Because the model postdates these games and held-out performance remains untested, this result should be interpreted as saturation of the public set only.
Chinese Translation
我们之前的ARC-AGI-3代理结合了可执行的世界建模、计划简化和精确重放验证，但未明确哪种思想对其性能起到了关键作用。我们通过四个嵌套的基于Codex的代理来解决这一归因问题：一个文本基准；一个没有重放验证的灵活接口可执行世界模型；同样的可执行模型加上计划简化；以及一个固定接口的验证处理，保留简化并要求精确再现记录的观察。主要研究评估了所有四个代理在公共ARC-AGI-3游戏中使用gpt-5.4和gpt-5.5进行高和超高推理努力的表现。探索性后续研究使用gpt-5.6-sol在超高和最大推理努力下评估文本和验证变体。最稳健的结果是，每个代理变体在更强的模型和更大的推理努力下都有所改善。在每个模型-努力设置中，各变体之间的差异小于预期，而各个组件的效果在不同设置中有所不同。要求持久的可执行交付物并不总是有利：在gpt-5.5的两个设置中，文本变体的表现优于灵活接口可执行变体。简化在四个模型-努力设置中的三个中提高了性能，唯一的例外是最弱的设置。完整的验证处理在所有四个设置中排名第一，尽管它使用了更多的资源。在gpt-5.6-sol后续研究中，验证变体在两个推理努力下完全解决了每个公共游戏，达到了约99%的RHAE，并且使用的总动作数不到人类基准的一半。由于该模型在这些游戏之后产生，且保留的性能尚未经过测试，因此该结果应被解读为公共集的饱和。
cs.AI / 8 / 2607.15442
Beyond a Joke: Multi-Angle Reasoning for Detecting and Explaining Harmful Humor in Memes
超越笑话：多角度推理用于检测和解释表情包中的有害幽默
Abstract
Internet memes intertwine visual cues, textual content, and cultural context, making them particularly challenging to interpret in scenarios where humor, sarcasm, and harmful intent coexist. These complexities highlight the need for explainable meme understanding systems that can provide reliable and structured reasoning to support both accurate classification and human interpretability. However, existing multimodal classifiers either overlook these interdependencies or provide only limited interpretability. In this paper, we introduce MAR-12, a novel framework that leverages Vision Language Models (VLMs) for meme detection and understanding in settings where humorous and hateful elements may coexist. The framework first interprets each meme through twelve structured perspectives derived from humor and hate theories. It then applies a role-aware soft-gated attention mechanism to learn how much each perspective should contribute, followed by a prototype-based classifier for the final prediction. Finally, explanations are synthesized using both perspective-specific reasoning and learned attention weights, ensuring transparent and context-grounded justifications. We evaluate MAR-12 on the PrideMM and Memotion datasets, where it achieves up to 80.3% accuracy for humor detection and 75.9% accuracy for hate detection, outperforming state-of-the-art approaches. Furthermore, both human and GPT-4-based evaluations confirm that MAR-12 produces coherent and persuasive explanations, particularly for memes in which humorous and harmful cues co-occur.
Chinese Translation
互联网表情包将视觉线索、文本内容和文化背景交织在一起，使得在幽默、讽刺和有害意图共存的情况下特别难以解读。这些复杂性突显了需要可解释的表情包理解系统，以提供可靠且结构化的推理，支持准确分类和人类可解释性。然而，现有的多模态分类器要么忽视了这些相互依赖性，要么仅提供有限的可解释性。本文介绍了MAR-12，一个新颖的框架，利用视觉语言模型（Vision Language Models, VLMs）在幽默和仇恨元素可能共存的环境中进行表情包检测和理解。该框架首先通过十二个基于幽默和仇恨理论的结构化视角来解读每个表情包。然后，它应用一种角色感知的软门控注意机制，以学习每个视角应贡献的程度，最后使用基于原型的分类器进行最终预测。最后，利用视角特定的推理和学习到的注意权重合成解释，确保透明且基于上下文的论证。我们在PrideMM和Memotion数据集上评估了MAR-12，其幽默检测准确率高达80.3%，仇恨检测准确率为75.9%，超越了最先进的方法。此外，基于人类和GPT-4的评估确认MAR-12生成了一致且有说服力的解释，特别是对于幽默和有害线索共存的表情包。
cs.AI / 9 / 2607.15459
From Black Box to Executable Logic: Explainable Reinforcement Learning through Prolog Expert Systems
从黑箱到可执行逻辑：通过 Prolog 专家系统实现可解释的强化学习
Abstract
A trained deep reinforcement learning policy is a black box, and we ask whether it can be made explainable by rewriting it as an executable logic program that reproduces its behaviour and that a person can read, a logic engine can run, and an optimizer can edit. We present a three-stage post-hoc transformation that extracts a frozen proximal policy optimization teacher, induces an ordered rule list from its decisions in the manner of classical relational learning, and emits the result as a Prolog program whose every decision is executed by an off-the-shelf logic engine; a subsequent expansion stage edits the rule base and accepts an edit only when policy evaluation certifies a return increase. We prove four guarantees. A return-loss bound makes the distilled program a machine-checkable certificate in a finite Markov decision process, and the expansion loop improves monotonically and terminates. For the continuous-observation setting we answer whether the conversion is possible at all: the propositional threshold instantiation converts the network to arbitrary fidelity as the resolution B grows, with disagreement O(1/B) and a return gap that closes at the same rate, and a matching lower bound shows the cost is exponential in the observation dimension for an oblique decision boundary. Empirically, on a two-room key-and-door task with 16,944 reachable states the expanded Prolog program attains exact optimal return in every seed and, in a budget-capped regime, exceeds the stochastic teacher on exact return in ten of ten seeds. On three continuous-control tasks the emitted program substitutes the network, matching the neural teacher within noise on Acrobot with eleven clauses and recovering about 97% of its return on CartPole, while on the finer-control LunarLander it recovers only partially, exactly the ceiling the exponential lower bound predicts.
Chinese Translation
训练好的深度强化学习策略是一个黑箱，我们探讨是否可以通过将其重写为一个可执行的逻辑程序，使其可解释，该程序能够再现其行为，并且可以被人类阅读、逻辑引擎运行和优化器编辑。我们提出了一种三阶段的后处理转换，提取一个冻结的近端策略优化教师，从其决策中诱导出一个有序规则列表，采用经典关系学习的方式，并将结果输出为一个 Prolog 程序，其每个决策都由现成的逻辑引擎执行；随后扩展阶段编辑规则库，仅在策略评估认证回报增加时接受编辑。我们证明了四个保证。回报损失界限使得提炼出的程序在有限马尔可夫决策过程中成为一个机器可验证的证书，而扩展循环单调改进并终止。对于连续观察设置，我们回答了转换是否可能：命题阈值实例化将网络转换为任意精度，随着解析度 B 的增加，分歧为 O(1/B)，回报差距以相同的速度收敛，而匹配的下界表明，对于倾斜决策边界，成本在观察维度上是指数级的。在经验上，在一个具有 16,944 个可达状态的两室钥匙与门任务中，扩展的 Prolog 程序在每个种子中都达到了精确的最优回报，并且在预算限制的情况下，在十个种子中超过了随机教师的精确回报。在三个连续控制任务中，发出的程序替代了网络，在 Acrobot 上与神经教师在噪声中匹配，使用了十一条子句，并在 CartPole 上恢复了约 97% 的回报，而在更精细控制的 LunarLander 上则仅部分恢复，正好是指数下界预测的上限。
cs.AI / 10 / 2607.15480
A Critical Analysis of Trustworthy AI Tools, Mark Frameworks, and the Implementation Chasms
对可信赖人工智能工具、信任标记框架及实施鸿沟的批判性分析
Abstract
As artificial intelligence (AI) systems increasingly impact society, ensuring their ethical and trustworthy deployment has become a global priority. While a myriad of high-level ethical guidelines have emerged, criticism persists that these frameworks remain abstract and lack concrete mechanisms for implementation. This paper conducts a critical analysis of tools and trust mark frameworks intended to operationalize trustworthy AI (TAI), drawing on a comprehensive dataset from the OECD. Through empirical mapping and descriptive comparative analysis, we identify significant asymmetries in ethical focus, lifecycle coverage, stakeholder targeting, and tool typology. Our findings show a strong emphasis on fairness, transparency, and robustness, with comparatively little attention paid to explainability, digital security, and environmental sustainability. Moreover, most tools and certifications concentrate on post-development stages, with limited guidance for early design or data collection phases. Educational initiatives and policy engagement are notably underdeveloped, suggesting that current TAI efforts are dominated by technical and procedural measures within industry contexts. We argue that bridging the persistent chasm between AI principles and practice requires expanding ethical objectives, embedding ethics across the AI lifecycle, and fostering broader multi-stakeholder participation. This study provides both a diagnosis of existing implementation gaps and actionable recommendations for advancing more holistic, inclusive, and enforceable AI governance
Chinese Translation
随着人工智能（AI）系统对社会的影响日益加深，确保其伦理和可信赖的部署已成为全球优先事项。尽管出现了大量高层次的伦理指导方针，但批评意见依然存在，认为这些框架过于抽象，缺乏具体的实施机制。本文对旨在实现可信赖人工智能（TAI）的工具和信任标记框架进行了批判性分析，基于来自经济合作与发展组织（OECD）的综合数据集。通过实证映射和描述性比较分析，我们识别出在伦理关注、生命周期覆盖、利益相关者定位和工具类型方面存在显著的不对称性。我们的研究结果显示，公平性、透明性和稳健性受到强烈关注，而对可解释性、数字安全和环境可持续性的关注相对较少。此外，大多数工具和认证集中于开发后的阶段，对早期设计或数据收集阶段的指导有限。教育倡议和政策参与明显不足，这表明当前的TAI努力主要受行业背景下的技术和程序措施主导。我们认为，弥合AI原则与实践之间持续存在的鸿沟需要扩展伦理目标，将伦理嵌入AI生命周期的各个环节，并促进更广泛的多利益相关者参与。本研究不仅诊断了现有实施差距，还提出了推动更全面、包容和可执行的AI治理的可行建议。
cs.AI / 11 / 2607.15532
Logic, Optimization, and Artificial Intelligence
逻辑、优化与人工智能
Abstract
Logic and optimization can, in combination, make valuable contributions to rule-based AI. Logic is the obvious medium for encoding a rule base and drawing inferences from it, while optimization provides a powerful technology for computing inferences. Their combination has taken on new relevance amid a growing concern for transparency in AI. which is important for reproducibility, explainability, trustworthiness, and fairness. Rule-based AI provides a natural solution to transparency that is becoming increasingly practical due to today's highly advanced optimization methods. This article surveys several areas of logic-optimization partnership, including probabilistic logic, Bayesian logic, belief logics and Dempster-Shafer theory, nonmonotonic (default) logic, many-valued logics, and inference of logical formulas from noisy data based on Boolean regression. It shows how to compute projections, the fundamental problem of both logic and optimization, using decision diagrams and logic-based Benders decomposition. It describes the use of postoptimality analysis to explain how conclusions are reached, further enhancing transparency, as well as the role of optimization in answer set programming modulo theories. The paper concludes by suggesting possible future research directions.
Chinese Translation
逻辑与优化的结合可以为基于规则的人工智能（AI）做出重要贡献。逻辑显然是编码规则库和从中推导推论的媒介，而优化则为推导计算提供了强大的技术支持。在对AI透明度日益关注的背景下，这种结合变得尤为重要，透明度对于可重复性、可解释性、可信性和公平性至关重要。基于规则的AI为透明度提供了一种自然的解决方案，随着当今高度先进的优化方法的出现，这种解决方案变得越来越可行。本文调查了逻辑与优化合作的多个领域，包括概率逻辑、贝叶斯逻辑、信念逻辑与邓普斯特-谢弗理论、非单调（默认）逻辑、多值逻辑，以及基于布尔回归从噪声数据中推导逻辑公式的推理。文章展示了如何使用决策图和基于逻辑的Benders分解来计算投影，这是逻辑和优化的基本问题。它描述了后最优性分析的应用，以解释如何得出结论，从而进一步增强透明度，以及优化在模态理论的答案集编程中的作用。最后，论文建议了未来可能的研究方向。
cs.AI / 12 / 2607.15550
SeerGuard: A Safety Framework for Mobile GUI Agents via World Model Prediction
SeerGuard：通过世界模型预测的移动图形用户界面代理安全框架
Abstract
Mobile graphical user interface (GUI) agents have demonstrated remarkable capabilities in automating complex tasks, yet they introduce critical safety risks where a single erroneous action can lead to irreversible consequences. Existing safety mechanisms are primarily reactive, lacking the ability to assess risks before execution. In this paper, we introduce SeerGuard, a consequence-aware safety framework designed to mitigate these risks through pre-execution instruction-level screening and action-level risk assessment. Specifically, the action-level assessment analyzes agent-proposed actions within current GUI states, anticipating likely outcomes to identify risks before they are executed. To enable these capabilities, we construct a unified safety-augmented world model (SAWM) via multi-task learning, integrating semantic next-state prediction with safety risk assessment. Extensive experiments demonstrate that SeerGuard generalizes effectively across diverse mobile GUI agents. On Qwen3-VL-8B-Instruct, it increases the safety-utility score from $0.191$ to $0.596$ at $\omega=0.8$ and reduces the risk-cost score from $0.347$ to $0.130$ at $\alpha=0.8$. Further analyses on our SAWM validate the effectiveness of the instruction-level screening, alongside the capability of action risk assessment and next-state prediction.
Chinese Translation
移动图形用户界面（GUI）代理在自动化复杂任务方面展现了显著的能力，但它们也引入了关键的安全风险，其中单个错误操作可能导致不可逆的后果。现有的安全机制主要是反应性的，缺乏在执行前评估风险的能力。本文提出了SeerGuard，一种关注后果的安全框架，旨在通过执行前的指令级筛选和动作级风险评估来减轻这些风险。具体而言，动作级评估分析代理在当前GUI状态下提出的动作，预测可能的结果，以在执行前识别风险。为了实现这些功能，我们通过多任务学习构建了一个统一的安全增强世界模型（SAWM），将语义下一个状态预测与安全风险评估相结合。大量实验表明，SeerGuard在多种移动GUI代理中有效泛化。在Qwen3-VL-8B-Instruct上，它将安全效用评分从$0.191$提高到$0.596$（$eta=0.8$），并将风险成本评分从$0.347$降低到$0.130$（$eta=0.8$）。对我们的SAWM的进一步分析验证了指令级筛选的有效性，以及动作风险评估和下一个状态预测的能力。
cs.AI / 13 / 2607.15592
MGDT: MLLM-Guided Diffusion Transformer with Relation-Adaptive Mixture-of-Experts for Multimodal Knowledge Graph Completion
MGDT：基于MLLM引导的关系自适应专家混合模型的扩散变换器用于多模态知识图谱补全
Abstract
Multimodal Knowledge Graph Completion (MKGC) requires inferring missing entities from structural, textual, and visual cues. Existing diffusion-based MKGC methods usually denoise directly on raw multimodal features. Such a design forces the denoiser to simultaneously perform relation-dependent cue selection, cross-modal semantic alignment, and structure-aware entity generation, which introduces noisy and semantically inconsistent conditions for diffusion and consequently leads to suboptimal completion performance. To address this limitation, we propose MGDT: MLLM-Guided Diffusion Transformer with Relation-Adaptive Mixture-of-Experts (MGDT), a novel MKGC framework built on an align-then-diffuse paradigm. MGDT first employs a Relation-Adaptive Semantic Routing Mixture-of-Experts (RASR-MoE) module to select relation-relevant multimodal semantic transformation paths and suppress irrelevant modality interference. MGDT then uses a frozen Multimodal Large Language Model (MLLM) as a semantic anchor to align the routed multimodal representations into a unified latent space and reduce cross-modal semantic heterogeneity. Finally, a Knowledge Graph Diffusion Transformer (KGDT) performs graph-conditioned denoising generation in the aligned space to produce the missing entity representation. Experiments on three benchmark datasets show that MGDT consistently outperforms strong baselines.
Chinese Translation
多模态知识图谱补全（MKGC）需要从结构、文本和视觉线索中推断缺失的实体。现有的基于扩散的MKGC方法通常直接对原始多模态特征进行去噪。这种设计迫使去噪器同时执行依赖关系的线索选择、跨模态语义对齐和结构感知的实体生成，这为扩散引入了噪声和语义不一致的条件，从而导致次优的补全性能。为了解决这一局限性，我们提出了MGDT：基于关系自适应专家混合模型的MLLM引导的扩散变换器（MGDT），这是一个基于对齐-再扩散范式的新型MKGC框架。MGDT首先采用关系自适应语义路由专家混合模型（RASR-MoE）模块选择与关系相关的多模态语义转换路径，并抑制无关模态的干扰。然后，MGDT使用一个冻结的多模态大型语言模型（MLLM）作为语义锚，将路由的多模态表示对齐到统一的潜在空间中，并减少跨模态语义异质性。最后，知识图谱扩散变换器（KGDT）在对齐空间中执行图条件去噪生成，以产生缺失的实体表示。在三个基准数据集上的实验表明，MGDT始终优于强基线。
cs.AI / 14 / 2607.15647
Neuro-Symbolic AI for LEED compliance: Document-Centric Benchmarking, Deterministic Numeric Checking, and When Multimodal Hurts
用于LEED合规的神经符号人工智能：以文档为中心的基准测试、确定性数值检查以及多模态的弊端
Abstract
LEED v4.1 BD+C certification remains a document-intensive process that requires reviewers to read hundreds of pages of project evidence and apply credit-specific threshold logic by hand. This paper investigates whether small, locally deployed language models can perform meaningful screening of LEED documentation and how deterministic symbolic components should share that work. A neuro-symbolic pipeline is introduced that aligns project PDFs to LEED credit sections, retrieves evidence with credit-aware keyword signatures, verifies compliance with a locally hosted 4-billion-parameter language model, and applies a LEED-specific numeric checker to quantitative thresholds. Experiments on four university buildings (484 PDFs, 153 credit-level decisions) show that a 4-billion-parameter model (gemma3:4b) is the strongest text-only core verifier, achieving 67.3% accuracy and outperforming a larger 8-billion-parameter model (llama3.1:8b) in this task. The deterministic numeric checker corrects arithmetic errors on key quantitative credits, moving EA-p2 from 50% to 100% accuracy and improving several other credits when required values are reliably extracted. At the same time, the full neuro-symbolic configuration achieves 61.6% overall accuracy, trailing the best text-only baseline due to extraction failures and conservative behavior on qualitative categories. Systematic ablations show that adding low-resolution drawing images (150-300 dpi) consistently reduces accuracy, and that prompt effectiveness depends on the building's ground-truth PASS rate: rubric prompts perform best on documentation-rich projects, while chain-of-thought prompts perform best on documentation-lean projects. Within the specific scope of LEED v4.1 BD+C compliance verification over raw project documentation, this pipeline and its baselines provide an initial reproducible reference point for both accuracy and failure modes.
Chinese Translation
LEED v4.1 BD+C 认证仍然是一个文档密集型的过程，要求审查员手动阅读数百页的项目证据并应用特定信用阈值逻辑。本文探讨了小型本地部署语言模型是否能够对LEED文档进行有意义的筛查，以及确定性符号组件应如何分担这项工作。我们提出了一种神经符号管道，该管道将项目PDF与LEED信用部分对齐，通过信用感知的关键词签名检索证据，利用本地托管的40亿参数语言模型验证合规性，并对定量阈值应用LEED特定的数值检查器。在对四座大学建筑（484个PDF，153个信用级别决策）的实验中，发现40亿参数模型（gemma3:4b）是最强的仅文本核心验证器，准确率达到67.3%，在这一任务中优于更大的80亿参数模型（llama3.1:8b）。确定性数值检查器纠正了关键定量信用的算术错误，将EA-p2的准确率从50%提升至100%，并在可靠提取所需值时改善了其他几个信用。同时，完整的神经符号配置总体准确率为61.6%，由于提取失败和在定性类别上的保守行为，落后于最佳的仅文本基线。系统的消融实验表明，添加低分辨率的绘图图像（150-300 dpi）会持续降低准确率，而提示的有效性取决于建筑的真实通过率：评分提示在文档丰富的项目上表现最佳，而思维链提示在文档稀缺的项目上表现最佳。在LEED v4.1 BD+C合规验证的特定范围内，该管道及其基线为准确性和失败模式提供了初步可重复的参考点。
cs.AI / 15 / 2607.15660
ToolVerse: Unlocking Massive Environments and Long-Horizon Tasks for Agentic Reinforcement Learning
ToolVerse：为自主强化学习解锁大规模环境和长时间任务
Abstract
While LLM agents demonstrate strong reasoning abilities in compact and well-defined scenarios, they struggle to maintain robustness and effectiveness when faced with large-scale, diverse, and dynamic real-world environments that demand seamless tool integration. To address this gap, we introduce ToolVerse, a comprehensive framework that scales up agentic RL environments and enables agents to perform complex long-horizon reasoning in Tool-Integrated Reasoning (TIR) tasks. First, ToolVerse automatically builds the massive executable agent training environments from nearly 400 real-world Model Context Protocols (MCPs) that contain about 4500 tools. Second, we propose a task design strategy based on a tool dependency graph, utilizing Dynamic Unlocking Sampling Algorithm to generate long-horizon tasks, and produce GUST (Graph Unlocking Sampling Tasks) dataset. Third, to alleviate the credit assigment problem in long-horizon agentic RL, we propose a fine-grained Turn-Aware Relative Advantage algorithm. We conduct extensive Agentic RL training using ToolVerse and evaluate our framework on serveral agentic benchmarks. Experimental results demonstrate that our framework significantly strengthens LLMs' capabilities in long-horizon tool use, achieving a marked performance boost and showcasing robust reasoning within dynamic environments.
Chinese Translation
尽管大型语言模型（LLM）代理在紧凑且定义明确的场景中表现出强大的推理能力，但在面对需要无缝工具集成的大规模、多样化和动态的现实世界环境时，它们在稳健性和有效性方面却显得力不从心。为了解决这一问题，我们提出了ToolVerse，一个全面的框架，旨在扩展自主强化学习环境，并使代理能够在工具集成推理（Tool-Integrated Reasoning, TIR）任务中进行复杂的长时间推理。首先，ToolVerse自动构建了来自近400个现实世界模型上下文协议（Model Context Protocols, MCPs）的庞大可执行代理训练环境，这些协议包含约4500个工具。其次，我们提出了一种基于工具依赖图的任务设计策略，利用动态解锁采样算法（Dynamic Unlocking Sampling Algorithm）生成长时间任务，并产生GUST（图解锁采样任务）数据集。第三，为了缓解长时间自主强化学习中的信用分配问题，我们提出了一种细粒度的转向感知相对优势算法（Turn-Aware Relative Advantage）。我们使用ToolVerse进行了广泛的自主强化学习训练，并在多个自主基准上评估了我们的框架。实验结果表明，我们的框架显著增强了LLM在长时间工具使用中的能力，实现了显著的性能提升，并展示了在动态环境中的稳健推理能力。
cs.AI / 16 / 2607.15686
S1-Omni: A Unified Multimodal Reasoning Model for Scientific Understanding, Prediction, and Generation
S1-Omni：一个统一的多模态推理模型用于科学理解、预测和生成
Zhao, Jiahao, Liu, Junyi, Xu, Lifeng, Xu, Nan, Wang, Qingli, Li, Qingxiao, Chen, Tianle, Wu, Xiaoyu, Zheng, Yawen, Wang, Zikai, Liu, Guanming, Zhou, Hequn, Wang, Jingyi, Shu, Jingyuan, Wang, Keqi, He, Li, Diao, Songyang, Xu, Wenhui, Ren, Xinyu, Fan, Yaqin, Zhou, Yujin, Yao, Zhanao
Abstract
We present S1-Omni, a unified multimodal reasoning model for scientific understanding, prediction, and generation. AI for Science (AI4S) has advanced significantly through domain-specific models, tool-augmented LLMs, and scientific language models. However, model capabilities remain highly fragmented, limiting the joint modeling of heterogeneous data, scientific laws, and expert knowledge. S1-Omni addresses this gap by consolidating these capabilities into a single, coherent scientific reasoning model. The architecture of S1-Omni is built upon three core components: unified representation of scientific data, natural-world knowledge alignment, and decoding for domain-specific tasks. First, S1-Omni maps natural-language instructions and scientific objects, including CIF, SMILES, protein sequences, spectra, and scientific images, into a shared representation space. Second, it incorporates scientific laws and expert knowledge into data construction and training, enabling the model to reason from scientific evidence. Third, it performs task-specific decoding to support a broad range of applications, including property prediction, spectrum-to-molecular generation, protein site and structure prediction, and scientific image generation and editing. S1-Omni is trained on S1-Omni-Corpus, which covers 200 scientific tasks and contains millions of reasoning samples, and is evaluated on over 60 scientific benchmarks. It outperforms GPT-5.5 and Gemini-3.1-Pro on most benchmarks and matches or surpasses domain-specific models on several benchmarks. Overall, S1-Omni provides a practical path toward unified scientific modeling.
Chinese Translation
我们提出了S1-Omni，一个用于科学理解、预测和生成的统一多模态推理模型。科学人工智能（AI for Science, AI4S）通过领域特定模型、工具增强的大型语言模型（LLMs）和科学语言模型取得了显著进展。然而，模型能力仍然高度碎片化，限制了异构数据、科学法则和专家知识的联合建模。S1-Omni通过将这些能力整合为一个单一、连贯的科学推理模型来填补这一空白。S1-Omni的架构基于三个核心组件：科学数据的统一表示、自然世界知识的对齐，以及针对特定领域任务的解码。首先，S1-Omni将自然语言指令和科学对象（包括CIF、SMILES、蛋白质序列、光谱和科学图像）映射到一个共享的表示空间。其次，它将科学法则和专家知识纳入数据构建和训练中，使模型能够基于科学证据进行推理。第三，它执行特定任务的解码，以支持广泛的应用，包括性质预测、光谱到分子的生成、蛋白质位点和结构预测，以及科学图像的生成和编辑。S1-Omni在S1-Omni-Corpus上进行训练，该数据集涵盖200个科学任务，包含数百万个推理样本，并在超过60个科学基准上进行评估。在大多数基准测试中，它的表现优于GPT-5.5和Gemini-3.1-Pro，并在多个基准上与领域特定模型相匹配或超越。总体而言，S1-Omni为统一科学建模提供了一条切实可行的路径。
cs.AI / 17 / 2607.15715
Behavioral Controllability of Agentic Models for Information Extraction: From Fixed Workflows to Reflective Agents
信息提取中代理模型的行为可控性：从固定工作流程到反思代理
Abstract
Large language model (LLM) agents are increasingly used for complex information-extraction tasks, yet it remains unclear whether agentic components such as reflection and memory lead to observable and controllable improvements over fixed LLM workflows. We study this question through conference-paper dataset extraction, where a system must identify datasets mentioned in scholarly PDFs and produce structured records. We compare a fixed workflow baseline with reflective agent variants and specify an optimized agent condition (S2) that extends the same task with richer PDF tools and dynamic tool selection. Our evaluation emphasizes process-level behavior--including tool execution, retries, reflection, memory use, runtime, and failure recovery--while treating extraction coverage and field completeness as secondary outcome measures. The paper characterizes when agentic mechanisms change system behavior, whether these changes improve task completion, and how the observed failure modes motivate an optimized agent design under the same evaluation harness.
Chinese Translation
大型语言模型（LLM）代理在复杂信息提取任务中的应用日益增多，但尚不清楚反思和记忆等代理组件是否能在固定的LLM工作流程上带来可观察和可控的改进。我们通过会议论文数据集提取来研究这个问题，其中系统必须识别学术PDF中提到的数据集并生成结构化记录。我们将固定工作流程基线与反思代理变体进行比较，并指定了一种优化代理条件（S2），该条件扩展了相同任务，提供了更丰富的PDF工具和动态工具选择。我们的评估强调了过程级行为，包括工具执行、重试、反思、记忆使用、运行时间和故障恢复，同时将提取覆盖率和领域完整性视为次要结果指标。本文描述了代理机制何时改变系统行为，这些变化是否改善了任务完成情况，以及观察到的故障模式如何促使在相同评估框架下优化代理设计。
cs.AI / 18 / 2607.15776
NeurOWL: An LLM-Based Neural-symbolic Framework for Incomplete OWL Ontology Reasoning
NeurOWL：基于大型语言模型的神经符号框架用于不完整OWL本体推理
Abstract
OWL ontologies provide a formal knowledge representation framework that enables semantic reasoning, and have been widely adopted across domains such as healthcare and bioinformatics. In practice, however, real-world ontologies are often incomplete, which pose challenges for reasoning. In this work, we focus on a fundamental subsumption reasoning problem: given an incomplete ontology and a candidate (non-entailed) subsumption, determine whether the subsumption is semantically plausible and, if so, providing a logically sound explanation containing potential missing axioms. This task unifies subsumption verification with ontology abduction, and generalizes the latter by removing the need for a predefined candidate set of missing axioms. To address this subsumption reasoning problem, we propose NeurOWL, an end-to-end neuro-symbolic framework that jointly performs verification and abduction, leveraging both formally defined semantics and textual semantics through Large Language Models and ontology embeddings. We evaluate NeurOWL on real-world ontologies across multiple domains, demonstrating strong and robust performance across different domains.
Chinese Translation
OWL本体提供了一种正式的知识表示框架，能够实现语义推理，并已在医疗和生物信息学等多个领域广泛应用。然而，在实际应用中，现实世界的本体往往是不完整的，这给推理带来了挑战。在本研究中，我们关注一个基本的子类推理问题：给定一个不完整的本体和一个候选（非蕴含的）子类，确定该子类在语义上是否合理，如果合理，则提供一个包含潜在缺失公理的逻辑上合理的解释。该任务将子类验证与本体归纳统一起来，并通过消除对预定义缺失公理候选集的需求来推广后者。为了解决这一子类推理问题，我们提出了NeurOWL，一个端到端的神经符号框架，能够联合执行验证和归纳，利用大型语言模型和本体嵌入，结合形式定义的语义和文本语义。我们在多个领域的现实世界本体上评估了NeurOWL，展示了其在不同领域的强大和稳健的性能。
cs.AI / 19 / 2607.15781
AgentFAIR: A Multi-Agent Collaborative Framework for FAIRness Evaluation of Geospatial Datasets
AgentFAIR：一个用于地理空间数据集公平性评估的多智能体协作框架
Abstract
Geospatial datasets support applications from urban planning to climate modeling, yet consistent assessment of FAIR compliance is difficult. Existing evaluators use different rubrics and evidence sources and may fail on JavaScript-rendered pages or repository-specific identifiers. For 50 datasets from 10 repositories, the standard deviation of normalized scores across available tools averages 15.0 percentage points and reaches 30.3 for one dataset. Because these outputs are not equivalent measurements, we use them to characterize disagreement and failure modes, not comparative accuracy. We present AgentFAIR, a multi-agent framework combining structured metadata extraction with 13 sub-principle-specific LLM evaluators. Each produces a 0-3 maturity score, cited evidence, and recommendations; a critic checks evidence and consistency and can request targeted re-evaluation. Mean Findability, Accessibility, Interoperability, and Reusability scores are 79.7%, 70.4%, 45.3%, and 72.0%. Rank correlations with four baseline tools range from 0.31 to 0.61; the FAIR-enough comparison is not statistically significant. On a 10-dataset repeated-run subset, sub-principle agreement averages 89% (standard deviation: 3 percentage points), versus 71% without the critic. A preliminary 15-dataset expert study yields Fleiss' kappa of 0.71 and 82% alignment with expert consensus. API cost is approximately USD 0.054 per dataset. These results support auditability and feasibility, while the limited benchmark, incomplete ablations, and single-model-family validation constrain claims about accuracy and generalization.
Chinese Translation
地理空间数据集支持从城市规划到气候建模的各种应用，但对 FAIR（可发现性、可获取性、互操作性和可重用性）合规性的持续评估却十分困难。现有评估工具使用不同的评分标准和证据来源，可能在 JavaScript 渲染页面或特定于存储库的标识符上失效。在来自 10 个存储库的 50 个数据集中，现有工具的标准化得分的标准差平均为 15.0 个百分点，其中一个数据集的标准差达到 30.3。由于这些输出并不是等效的测量，我们将其用于表征不一致性和失败模式，而非比较准确性。我们提出了 AgentFAIR，一个结合结构化元数据提取与 13 个子原则特定 LLM（大语言模型）评估器的多智能体框架。每个评估器生成一个 0-3 的成熟度评分、引用证据和建议；一个批评者检查证据和一致性，并可以请求针对性的重新评估。可发现性、可获取性、互操作性和可重用性的平均得分分别为 79.7%、70.4%、45.3% 和 72.0%。与四个基线工具的排名相关性范围为 0.31 到 0.61；公平性比较在统计上并不显著。在一个包含 10 个数据集的重复运行子集中，子原则一致性平均为 89%（标准差：3 个百分点），而没有批评者时为 71%。初步的 15 个数据集专家研究得出的 Fleiss' kappa 值为 0.71，与专家共识的对齐率为 82%。API 成本约为每个数据集 0.054 美元。这些结果支持可审计性和可行性，但有限的基准、未完成的消融实验和单一模型家族的验证限制了对准确性和泛化能力的主张。
cs.AI / 20 / 2607.15845
Knowledge-Centric Agents for Workflow Generation
以知识为中心的工作流生成代理
Abstract
Workflow generation in visual creation systems such as ComfyUI demands not only syntactic accuracy but also expert-level reasoning over modular compositions. Existing large language model (LLM) approaches often treat this as a direct text-to-JSON generation task, struggling with structural brittleness and lacking the experiential knowledge required for effective design. We argue that successful workflow generation requires modeling knowledge itself, including its structure, hierarchy, and reasoning dynamics. To this end, we propose a knowledge-centric framework that learns to invert, inject, and infer with knowledge across multiple abstraction levels. We first perform knowledge inversion to distill hierarchical representations, ranging from full pseudo-codes and skeletons to high-level strategies, from large collections of real-world workflows. We then conduct knowledge injection through supervised fine-tuning, teaching the model to reason from task descriptions to strategies and from strategies to executable structures. During inference, the model performs reversible reasoning to synthesize executable workflows, augmented by self-refinement for structural coherence. Extensive experiments demonstrate that our method produces workflows with richer node diversity, more coherent structures, and higher execution success rates than existing systems, establishing a new foundation for knowledge-driven, agentic workflow generation.
Chinese Translation
在 ComfyUI 等视觉创作系统中，工作流生成不仅需要语法准确性，还需要对模块化组合进行专家级推理。现有的大型语言模型（LLM）方法通常将其视为直接的文本到 JSON 生成任务，面临结构脆弱性的问题，并缺乏有效设计所需的经验知识。我们认为，成功的工作流生成需要对知识本身进行建模，包括其结构、层次和推理动态。为此，我们提出了一种以知识为中心的框架，学习在多个抽象层次上反转、注入和推理知识。我们首先进行知识反转，从大量真实世界工作流中提炼出层次化表示，涵盖完整的伪代码、框架到高层策略。然后通过监督微调进行知识注入，教会模型从任务描述推理到策略，再从策略推理到可执行结构。在推理过程中，模型执行可逆推理以合成可执行工作流，并通过自我优化增强结构一致性。大量实验表明，我们的方法生成的工作流在节点多样性、结构一致性和执行成功率方面优于现有系统，为知识驱动的代理工作流生成奠定了新的基础。
cs.AI / 21 / 2607.15901
DSWorld: A Data Science World Model for Efficient Autonomous Agents
DSWorld：一种用于高效自主智能体的数据科学世界模型
Abstract
Despite strong capabilities in data understanding and decision-making, autonomous data science agents still heavily rely on trial-and-error workflows that involve expensive computation. This bottleneck motivates models that can anticipate the effects of data science operations before real execution. In this paper, we introduce the concept of Data Science World Model, which model the data science execution environment by predicting environment state transitions conditioned on current workflow states and candidate operations. We further propose DSWorld, a practical framework that combines structured state construction, cost-aware routing, lightweight real execution, and an LLM-based simulator for expensive operations. To support training, we construct an 8K-scale transition trajectory dataset and introduce Reflective World Model Optimization, an error-aware reinforcement learning strategy for improving transition prediction. Experiments show that DSWorld accelerates RL-based agent training by approximately $14\times$ and search-based inference by approximately $3$-$6\times$ while maintaining competitive performance, and outperforms the strongest LLM baseline by 35.6% on transition prediction tasks. The code is available at https://anonymous.4open.science/r/DSWorld.
Chinese Translation
尽管自主数据科学智能体在数据理解和决策制定方面具备强大的能力，但它们仍然严重依赖于涉及高昂计算成本的试错工作流程。这一瓶颈促使我们开发能够在实际执行之前预测数据科学操作效果的模型。本文介绍了数据科学世界模型（Data Science World Model）的概念，该模型通过预测基于当前工作流程状态和候选操作的环境状态转变来模拟数据科学执行环境。我们进一步提出了DSWorld，这是一个实用框架，结合了结构化状态构建、成本感知路由、轻量级真实执行和基于大型语言模型（LLM）的昂贵操作模拟器。为了支持训练，我们构建了一个8K规模的转移轨迹数据集，并引入了反思世界模型优化（Reflective World Model Optimization），这是一种错误感知的强化学习策略，用于改善转移预测。实验表明，DSWorld将基于强化学习的智能体训练加速了约14倍，基于搜索的推理加速了约3至6倍，同时保持了竞争性能，并在转移预测任务中超越了最强的LLM基线35.6%。代码可在https://anonymous.4open.science/r/DSWorld获取。
cs.AI / 22 / 2607.15987
A Formally Grounded ODRL Evaluator: Implementation and Comparison
一个形式化基础的ODRL评估器：实现与比较
Abstract
The ODRL policy language is emerging as the de-facto standard for policy modelling data access and usage preferences, AI governance policies and data workflows in European dataspaces. The current standard has no mathematical formal semantics to describe how a system should implement policy evaluation. This has resulted in a variety of systems and tools that implement their own interpretation of the language, which limits interoperability and cannot guarantee consistent results. Based on an existing semantic model of ODRL, we formalise the problems of ODRL evaluation for the access control and monitoring scenarios, in both static and streaming settings, and we provide a novel, efficient algorithm and implementation. We present the first ODRL Evaluator with transparent formal semantics and supporting all rule types. We experimentally measure its performance, analysing different scalability dimensions related to policy complexity and size of the data on which a policy is evaluated. We compare our system with the state-of-the-art by providing a comparative review of existing ODRL evaluators, which highlights the differences in supported ODRL features and evaluation modes.
Chinese Translation
ODRL政策语言正在成为欧洲数据空间中数据访问和使用偏好、人工智能治理政策以及数据工作流建模的事实标准。目前的标准没有数学形式语义来描述系统应如何实现政策评估。这导致了多种系统和工具实现了各自对该语言的解释，从而限制了互操作性，并且无法保证结果的一致性。基于现有的ODRL语义模型，我们形式化了ODRL评估在访问控制和监控场景中的问题，包括静态和流式设置，并提供了一种新颖、高效的算法和实现。我们展示了第一个具有透明形式语义并支持所有规则类型的ODRL评估器。我们通过实验测量其性能，分析与政策复杂性和评估政策的数据大小相关的不同可扩展性维度。我们通过提供现有ODRL评估器的比较评审，将我们的系统与最先进的技术进行了比较，突出了支持的ODRL特性和评估模式的差异。
cs.AI / 23 / 2607.15992
Closing the AI Trust Gap: The Case for Independent Certification for Trustworthy AI
弥合人工智能信任缺口：独立认证可信赖人工智能的案例
Abstract
Over the past decade, responsible AI (RAI) has produced a substantial body of practice for identifying and mitigating the risks AI poses in high-stakes settings. Yet this work has not produced a market that rewards trustworthiness. Firms that invest seriously in safety, fairness, and oversight cannot consistently prove to consumers, regulators, and shareholders that their systems go beyond the bare minimum of compliance. What is missing is a way for society to recognize or compare the difference. The result is a trust gap: a structural condition in which responsible development efforts happen inside organizations but produce no external, independently recognized and verifiable signal of trustworthy outcomes. We argue this gap is sustained in part because of a focus on responsible AI (a matter of internal process) as opposed to trustworthy AI (a matter of independently verifiable real-world outcomes), and that it persists because of three compounding failures: (1) the market cannot distinguish trustworthy systems from their imitations; (2) evaluation targets models and outputs rather than deployed sociotechnical systems and their outcomes; (3) the measurement ecosystem is oriented toward avoiding harm rather than demonstrating benefit. Reviewing existing AI governance instruments and comparing them to certification regimes in healthcare, sustainability, and security, we show that none integrate a governance baseline, independently verified positive-outcome evidence, and market signaling in a single framework. We propose independent, outcome-oriented certification as the connective layer that can close the trust gap, complementing regulation and internal governance by making trustworthiness measurable, comparable, and commercially rewarded.
Chinese Translation
在过去十年中，负责任的人工智能（Responsible AI, RAI）为识别和减轻人工智能在高风险环境中所带来的风险产生了大量实践。然而，这项工作并未形成一个能够奖励可信赖性的市场。那些在安全、公平和监督方面认真投资的公司，无法持续向消费者、监管者和股东证明其系统超越了最低合规要求。缺失的是一种让社会能够识别或比较差异的方式。结果是形成了一个信任缺口：一种结构性条件，其中负责任的发展努力发生在组织内部，但未能产生外部的、独立认可且可验证的可信赖结果信号。我们认为，这一缺口部分是由于对负责任的人工智能（内部过程问题）的关注，而非可信赖的人工智能（独立可验证的现实结果问题），并且由于三种累积性失败而持续存在：（1）市场无法区分可信赖系统与其仿制品；（2）评估目标是模型和输出，而非已部署的社会技术系统及其结果；（3）测量生态系统的方向是避免伤害，而非展示利益。通过审查现有的人工智能治理工具，并将其与医疗、可持续性和安全领域的认证机制进行比较，我们发现没有一个框架能够将治理基线、独立验证的积极结果证据和市场信号整合在一起。我们建议采用独立的、以结果为导向的认证作为弥合信任缺口的连接层，补充监管和内部治理，使可信赖性可测量、可比较并获得商业奖励。
cs.AI / 24 / 2607.16038
SciForge: An AI-Native, Multimodal Workbench for Scientific Discovery
SciForge：一种原生人工智能的多模态科学发现工作台
Abstract
Scientific work increasingly spans heterogeneous artifacts -- papers, code, datasets, scientific file formats, model outputs, figures, manuscripts, and team decisions -- yet general-purpose AI assistants rarely preserve these objects as a coherent, auditable research state. We present SciForge, a multimodal research-native AI workbench that reserves the graphical interface for human judgment while search, parsing, model routing, workflow execution, plotting, writing, and presentation generation run as modular agent-accessible services. SciForge is built around five pillars: (i) \emph{goal-scoped scientific decision governance} for \textbf{goal-oriented} research, with review gates and shared review surfaces; (ii) \emph{translate-then-reason} for \textbf{multimodal} input, routing scientific objects through domain translators before the agent reasons; (iii) \emph{evidence governance} for \textbf{auditable} traceability, linking claims to provenance chains and audit findings; (iv) \emph{collaborative team science} for \textbf{collaborative} research, enabling multi-role decision governance, with shared team workspaces planned for future releases; and (v) \emph{real-world application scenarios} for \textbf{practical} impact, demonstrated through eight end-to-end user cases, with flagship demonstrations including multi-day agentic research sprints for gene discovery, AI-guided de novo protein design, molecular optimization, and genome-to-BGC discovery. The system combines a thin interaction layer, contextual research capability patterns, an Agent Runtime and Workflow Engine, an Evidence-DAG audit sidecar and a Scientific Model Router. SciForge currently runs as a desktop application, with mobile supervision support; future releases will deepen team collaboration. The system is open-source and available at https://github.com/AGI4Sci/SciForge
Chinese Translation
科学工作日益涉及异构的文献资料——论文、代码、数据集、科学文件格式、模型输出、图形、手稿和团队决策——然而，通用的人工智能助手很少能将这些对象作为一个连贯的、可审计的研究状态保存。我们提出了SciForge，一种多模态的研究原生人工智能工作台，它为人类判断保留了图形界面，同时搜索、解析、模型路由、工作流执行、绘图、写作和演示生成作为模块化的代理可访问服务运行。SciForge建立在五个支柱之上：（i）面向目标的科学决策治理，适用于目标导向的研究，设有审查门和共享审查界面；（ii）翻译后推理，适用于多模态输入，在代理推理之前通过领域翻译器路由科学对象；（iii）证据治理，确保可审计的可追溯性，将主张与来源链和审计发现关联；（iv）协作团队科学，适用于协作研究，支持多角色决策治理，并计划在未来版本中提供共享团队工作空间；（v）现实世界应用场景，确保实际影响，通过八个端到端用户案例进行演示，旗舰演示包括基因发现的多日代理研究冲刺、人工智能指导的从头蛋白设计、分子优化和基因组到BGC的发现。该系统结合了薄交互层、上下文研究能力模式、代理运行时和工作流引擎、证据有向无环图审计侧车以及科学模型路由器。SciForge目前作为桌面应用程序运行，并支持移动监督；未来版本将深化团队协作。该系统是开源的，可在https://github.com/AGI4Sci/SciForge获取。
cs.AI / 25 / 2607.16112
Harmonizing AI Safety Thresholds
协调人工智能安全阈值
Abstract
Frontier AI companies have published capability thresholds that differ substantially, making it difficult for third parties to verify whether a threshold has been crossed or to compare requirements across companies. Moreover, without common minimum thresholds, risk mitigation may be inconsistent, creating a potential race to the bottom in safety standards. We develop a methodology for deriving harmonized thresholds across three risk domains. For misuse risks (cyber and biological), we take expected harm as the key primitive and use an explicit risk-modeling approach that accounts for risk channels and model release conditions. For automated AI R&D, we base our proposed threshold on the observed rate of AI progress rather than expected harm. Our analysis expands upon prior work and highlights existing empirical gaps and limitations.
Chinese Translation
前沿人工智能公司发布的能力阈值差异显著，使得第三方难以验证是否已跨越阈值或比较各公司之间的要求。此外，缺乏共同的最低阈值可能导致风险缓解措施不一致，从而在安全标准上形成潜在的竞争下降。我们开发了一种在三个风险领域内推导协调阈值的方法论。对于误用风险（网络和生物），我们将预期伤害作为关键原始概念，并采用明确的风险建模方法，考虑风险渠道和模型发布条件。对于自动化人工智能研发，我们提出的阈值基于观察到的人工智能进展速度，而非预期伤害。我们的分析扩展了先前的研究，并突出了现有的实证空白和局限性。
cs.AI / 26 / 2607.16122
CRAFT: Clustering Rubrics to Diagnose Weak LLM Capabilities and Generate Targeted Fine-Tuning Data
CRAFT：聚类评分标准以诊断弱LLM能力并生成针对性的微调数据
Abstract
Evaluations should do more than measure a models current performance. They should tell us what to fix for the next model iteration and provide a way to generate targeted post training data. Most evaluation pipelines identify weak examples, topics, or categories, but they leave the underlying capability failure implicit: they say where a model fails, not why. We introduce CRAFT, a method that converts any rubric based evaluation dataset into a model specific diagnosis of weak capabilities. CRAFT treats each grading criterion as a capability probe: it extracts a capability description from every prompt rubric pair, clusters these descriptions into a hierarchical capability tree, scores the target model at every node, and selects low performing nodes dynamically across tree levels, at the granularity where each failure is clearest. The selected weak capabilities then direct the generation of targeted supervised finetuning data. Holding the data generation, finetuning, and evaluation setup fixed, we compare CRAFT against prompt level EvalTree clustering and untargeted random generation on four open source models, two professional domains (finance and legal), and 13 held out benchmarks disjoint from the diagnostic data. CRAFT achieves the strongest finance domain average for all four models under repeated temperature decoding; on legal domain, it is strongest for three of four models and remains within the decoding variance bands of the best baseline on the fourth. Diagnosing weaknesses at the level of rubric criteria, rather than prompts or categories, thus yields both a sharper picture of what a model cannot do and measurably better models after finetuning on that diagnosis.
Chinese Translation
评估不仅应衡量模型当前的性能，还应告诉我们在下一个模型迭代中需要修复的内容，并提供生成针对性后训练数据的方法。大多数评估流程识别出弱示例、主题或类别，但它们将潜在的能力失败隐含化：它们说明模型失败的地方，而不是失败的原因。我们提出CRAFT，一种将任何基于评分标准的评估数据集转换为特定模型的弱能力诊断的方法。CRAFT将每个评分标准视为能力探测器：它从每个提示和评分标准对中提取能力描述，将这些描述聚类成一个层次能力树，在每个节点对目标模型进行评分，并在每个失败最明显的粒度上动态选择低性能节点。所选的弱能力随后指导生成针对性的监督微调数据。在固定数据生成、微调和评估设置的情况下，我们将CRAFT与基于提示的EvalTree聚类和无目标随机生成进行了比较，涉及四个开源模型、两个专业领域（金融和法律）以及13个与诊断数据不重叠的基准。CRAFT在重复温度解码下为所有四个模型实现了金融领域的最佳平均表现；在法律领域，对于四个模型中的三个表现最佳，并且在第四个模型中仍然处于最佳基线的解码方差范围内。因此，在评分标准的层面上诊断弱点，而不是提示或类别，能够提供更清晰的模型无法执行的任务的图景，并在基于该诊断的微调后显著改善模型性能。
cs.CL / 1 / 2607.15380
Large Language Models as Unified Multimodal Learners for Clinical Prediction
大型语言模型作为统一的多模态学习者用于临床预测
Abstract
Electronic health records combine free-text clinical narratives with structured measurements such as vital signs, laboratory values, and comorbidities. Yet most clinical prediction systems still rely on task-specific fusion architectures, pairing dedicated encoders for each modality with learned combination mechanisms that must be re-engineered for every new task and clinical setting. We propose a simpler alternative: convert all patient data, regardless of modality, into a single natural language sequence and fine-tune a pretrained language model end-to-end, with no architectural modification for fusion. We evaluate this approach across three clinically distinct prediction tasks: in-hospital mortality on MIMIC-III, graft failure prediction using longitudinal data from a German transplant center, and emergency triage classification from ambulance records - comparing encoder-based (ModernBERT) and decoder-based (Llama 3.1, Gemma, DeepSeek-R1-Qwen, Qwen3) fine-tuning against established multimodal baselines and, for graft failure, a gradient boosting model currently used in clinical practice for post-transplant patient management. Across all three tasks, unified textual serialization matches or exceeds task-specific multimodal baselines, and outperforms the clinically deployed gradient boosting system on graft failure prediction. These results indicate that a single serialization-based paradigm, without bespoke fusion architectures, is sufficient for multimodal clinical prediction - substantially reducing system complexity while matching or exceeding specialized designs.
Chinese Translation
电子健康记录将自由文本的临床叙述与结构化的测量数据（如生命体征、实验室数值和合并症）相结合。然而，大多数临床预测系统仍依赖于特定任务的融合架构，为每种模态配备专用编码器，并使用学习的组合机制，这些机制必须为每个新任务和临床环境重新设计。我们提出了一种更简单的替代方案：将所有患者数据（无论模态如何）转换为单一的自然语言序列，并对预训练语言模型进行端到端的微调，无需进行融合的架构修改。我们在三个临床上不同的预测任务中评估了这种方法：在MIMIC-III上的院内死亡率、使用来自德国移植中心的纵向数据进行的移植物失败预测，以及从救护车记录中进行的紧急分诊分类——将基于编码器的微调（ModernBERT）与基于解码器的微调（Llama 3.1、Gemma、DeepSeek-R1-Qwen、Qwen3）与已建立的多模态基线进行比较，并且在移植物失败预测中与目前用于临床实践的梯度提升模型进行比较。在所有三个任务中，统一的文本序列化与特定任务的多模态基线相匹配或超越，并在移植物失败预测中优于临床应用的梯度提升系统。这些结果表明，基于单一序列化的范式，无需定制的融合架构，对于多模态临床预测是足够的——在显著降低系统复杂性的同时，匹配或超越专门设计的效果。
cs.CL / 2 / 2607.15495
Verbalizable Representations Form a Global Workspace in Language Models
可言表征在语言模型中形成全球工作空间
Abstract
Out of everything the human brain processes, only a small fraction is consciously accessible, in the sense of being available for verbal report, deliberate control, and flexible reasoning. In this paper, we present evidence that an analogous functional distinction has emerged in large language models. Using a new interpretability technique, the Jacobian lens, we identify the representations a model is poised to verbalize at any point in its processing. These representations, which we collectively call the J-space, exhibit the functional properties characteristic of a global workspace: their contents can be reported, deliberately summoned and held, used to carry the intermediate steps of silent reasoning, and passed as arguments to arbitrary downstream computations, while automatic processing such as text parsing and routine inference proceeds without them. The J-space also has structural signatures that global workspace theory associates with conscious access: it carries coherent content only in an intermediate band of layers, holds on the order of tens of concepts at a time, and is broadcast by the model's weights more widely than other representations. These properties make it a practical window into a model's unspoken thinking. In alignment audits, it reveals strategic deliberation, evaluation awareness, and trained-in misaligned dispositions that never appear in the model's outputs. We find that post-training installs the Assistant's point of view in the workspace, and we introduce counterfactual reflection training, which improves behavior by training only what a model would say if interrupted and asked to reflect. These results indicate that language models maintain a small, privileged set of representations bearing some of the functional hallmarks of conscious access, and that decoding these representations sheds light on ongoing cognitive processes.
Chinese Translation
在人类大脑处理的所有信息中，只有一小部分是可以被意识访问的，这意味着这些信息可以进行语言报告、故意控制和灵活推理。在本文中，我们提供了证据，表明在大型语言模型中出现了类似的功能区分。通过一种新的可解释性技术——雅可比透镜（Jacobian lens），我们识别出模型在其处理的任何时刻准备进行言语表达的表征。这些表征我们统称为 J-space，展现出全球工作空间特有的功能属性：其内容可以被报告、故意召唤和保持，用于承载无声推理的中间步骤，并作为参数传递给任意下游计算，而自动处理如文本解析和常规推理则在没有它们的情况下进行。J-space 还具有全球工作空间理论与意识访问相关的结构特征：它仅在中间层带中携带连贯内容，同时一次持有数十个概念，并且模型的权重比其他表征更广泛地广播这些内容。这些特性使其成为了解模型未言思维的实用窗口。在一致性审计中，它揭示了战略性思考、评估意识和训练中未对齐的倾向，这些在模型的输出中从未出现。我们发现，后训练阶段将助手的观点安装在工作空间中，并引入了反事实反思训练，通过仅训练模型在被打断并要求反思时会说的话来改善行为。这些结果表明，语言模型维持了一小组特权表征，具备某些意识访问的功能特征，并且解码这些表征可以揭示正在进行的认知过程。
cs.CL / 3 / 2607.15498
VarRate: Training-Free Variable-Rate KV Cache Compression for Long-Context LLMs
VarRate：无训练的可变速率键值缓存压缩用于长上下文大语言模型
Abstract
The key-value (KV) cache is the main memory bottleneck in long-context large language model (LLM) inference. Two leading training-free families are both structurally limited: token-selection methods (SnapKV, Ada-KV) score importance from an observation window and evict low-scoring tokens, but eviction is irreversible -- so when the importance signal degrades under query-agnostic reuse, accuracy collapses by 11-15 points; uniform low-rank coding keeps every token but spends equal rank everywhere, wasting budget. We observe that both failures share one cure: rank should be allocated, not evicted. We present VarRate, a training-free KV codec that assigns each token a variable low-rank budget by its query salience, keeping every token at a nonzero rank. Comparable adaptive-rank codecs reach this allocation only through training; VarRate requires none. Because no token is dropped, it degrades by only 3.5-5.5 points where query-aware selection collapses. At a matched 20% budget on LongBench (16 tasks), VarRate stays within 0.8 points of the uncompressed model on both Llama-3.1-8B and Qwen2.5-7B. Averaged over the two, it is the strongest matched-memory compressor. It significantly beats its uniform-rank ablation on both models. Against KVzip, a method purpose-built for query-agnostic reuse, it is accuracy-equivalent in three of four settings and within a point overall, at about one-eighth the prefill overhead.
Chinese Translation
键值（KV）缓存是长上下文大语言模型（LLM）推理中的主要内存瓶颈。两种主要的无训练方法在结构上都有局限性：令牌选择方法（SnapKV, Ada-KV）通过观察窗口评估重要性并驱逐低评分的令牌，但驱逐是不可逆的——因此，当重要性信号在查询无关重用下下降时，准确性会下降11-15个百分点；均匀低秩编码保留每个令牌，但在各处花费相同的秩，浪费预算。我们观察到这两种失败共享一个解决方案：秩应该被分配，而不是被驱逐。我们提出了VarRate，这是一种无训练的KV编码器，通过查询显著性为每个令牌分配可变的低秩预算，保持每个令牌的非零秩。可比的自适应秩编码器仅通过训练才能达到这种分配；而VarRate则不需要任何训练。由于没有令牌被丢弃，因此在查询感知选择崩溃的情况下，其准确性仅下降3.5-5.5个百分点。在LongBench（16个任务）上以匹配的20%预算，VarRate在Llama-3.1-8B和Qwen2.5-7B模型上与未压缩模型的差距保持在0.8个百分点以内。平均而言，它是最强的匹配内存压缩器。在这两个模型上，它显著优于均匀秩消融实验。与KVzip（专为查询无关重用而设计的方法）相比，在四种设置中的三种情况下，其准确性相当，总体上差距在一个百分点以内，且预填充开销约为其八分之一。
cs.CL / 4 / 2607.15544
EpiNarrate: Agentic Generation of Grounded Narratives from Epidemiological Scenario Projections
EpiNarrate：基于流行病学情景预测的自主生成扎根叙事
Abstract
Generation of clear and accessible public health narratives is critical for communicating complex epidemiological projections to policymakers and the general public at large. Such narratives require more than simply reporting numbers: projections must be contextualized and quantitatively grounded across multiple dimensions. Further, projections are often derived from large ensemble datasets which combine intervention assumptions, geographic and demographic strata, outcomes, time horizons, and uncertainty quantiles. However, directly using large language models (LLMs) to summarize and contextualize such data often leads to inconsistencies, omissions, and fragile behavior. We introduce an agentic framework (EpiNarrate) for public health report generation that separates structured numerical reasoning from natural-language generation. The framework first extracts scenario axes and organizes them into a partial-order schema, enabling systematic traversal of the underlying multidimensional space. It then constructs an augmented dataset and derives valid quantitative statements through a comparison grammar that enforces semantic and arithmetic consistency. To balance coverage and non-redundancy, we introduce an interestingness-driven selection mechanism based on maximum-entropy principles. Experiments on the COVID-19 Scenario Modeling Hub demonstrate that our model produces narratives with improved factual grounding and broader coverage of salient epidemiological patterns, while preserving the style of expert-written reports.
Chinese Translation
生成清晰且易于理解的公共卫生叙事对于向政策制定者和广大公众传达复杂的流行病学预测至关重要。这类叙事不仅仅是报告数字：预测必须在多个维度上进行背景化和定量扎根。此外，预测通常源自大型集合数据集，这些数据集结合了干预假设、地理和人口层次、结果、时间范围以及不确定性分位数。然而，直接使用大型语言模型（LLMs）来总结和背景化这些数据往往会导致不一致、遗漏和脆弱的行为。我们提出了一种自主框架（EpiNarrate）用于公共卫生报告生成，该框架将结构化的数值推理与自然语言生成分开。该框架首先提取情景轴并将其组织成部分有序模式，从而实现对基础多维空间的系统遍历。然后，它构建一个增强的数据集，并通过比较语法推导有效的定量陈述，该语法强制执行语义和算术一致性。为了平衡覆盖率和非冗余性，我们引入了一种基于最大熵原则的趣味驱动选择机制。在COVID-19情景建模中心的实验表明，我们的模型生成的叙事在事实基础和显著流行病学模式的覆盖面上有所改善，同时保持了专家撰写报告的风格。
cs.CL / 5 / 2607.15557
SkillCorpus: Consolidating and Evaluating the Open Skill Ecosystem for Real-World LLM Agents
SkillCorpus：整合与评估现实世界 LLM 代理的开放技能生态系统
Abstract
Agent skills, SKILL.md files that package reusable procedural knowledge for an LLM agent, are a popular mechanism for extending agent capabilities. Public repositories now host them in large and growing numbers, yet these artifacts are fragmented, redundant, and uneven in quality, and their value in practice is unclear. A core question remains open, namely how to consolidate this open-source SKILL.md ecosystem into a single usable corpus, and what bounds its benefit on real-world agent tasks. We present SkillCorpus, a framework that aggregates, curates, matches, and evaluates the open skill ecosystem at scale. It filters ~821,000 crawled skills through a multi-stage pipeline into 96,401 skills organised by a 16-class taxonomy and three quality facets (utility, robustness, safety), and pairs them with a fine-tuned retrieval-and-selection stack that matches task-relevant skills. We evaluate end-to-end across three benchmarks (SkillsBench, GDPVal, QwenClawBench), two harnesses, and two open backbones with a frontier robustness check. Integrating SkillCorpus yields consistent gains across all three benchmarks, largest on SkillsBench (+7.5 pp). An operational analysis traces the gains to a coverage boundary and a harness boundary. SkillCorpus is, to our knowledge, the first end-to-end account of when a curated, retrieval-served community corpus improves real agent tasks, and where it does not. The dataset, models, and code will be released upon acceptance.
Chinese Translation
代理技能，即为 LLM 代理打包可重用程序知识的 SKILL.md 文件，是扩展代理能力的一种流行机制。公共仓库中现在托管着大量不断增长的这些文件，然而这些文档存在碎片化、冗余和质量不均等问题，其在实践中的价值尚不明确。一个核心问题仍然悬而未决，即如何将这一开源的 SKILL.md 生态系统整合成一个可用的单一语料库，以及它在现实代理任务中的效益边界。我们提出了 SkillCorpus，这是一个在大规模上聚合、策划、匹配和评估开放技能生态系统的框架。它通过多阶段管道过滤约 821,000 个爬取的技能，最终整理出 96,401 个技能，按 16 类分类法和三个质量维度（实用性、稳健性、安全性）组织，并与一个经过微调的检索与选择堆栈相匹配，以匹配任务相关的技能。我们在三个基准（SkillsBench、GDPVal、QwenClawBench）、两个测试平台和两个开放骨干网络上进行了端到端评估，并进行了前沿稳健性检查。整合 SkillCorpus 在所有三个基准上均带来了持续的提升，其中在 SkillsBench 上的提升最大（+7.5 个百分点）。操作分析将这些提升归因于覆盖边界和测试平台边界。根据我们的知识，SkillCorpus 是首个全面阐述经过策划的、检索服务的社区语料库何时改善真实代理任务、何时未能改善的研究。数据集、模型和代码将在接受后发布。
cs.CL / 6 / 2607.15610
Process Reward Informed Tree Rollout for Effective Multi-Turn RL
基于过程奖励的信息树展开以实现有效的多回合强化学习
Abstract
Reinforcement learning (RL) has become a key approach for training LLM agents, yet popular methods such as GRPO/RLOO rely on multiple independently sampled complete trajectories for advantage estimation. In long-horizon agentic tasks, such a uniform rollout strategy can waste budget on uninformative dead-end attempts, while promising intermediate states do not receive sufficient exploration. The multi-turn structure of agentic trajectories, with interleaved actions and observations, naturally supports organizing a trajectory group as a tree, where each turn serves as a decision point for exploration. This perspective reframes effective exploration as the problem of deciding where to branch. We propose Process-Scorer Guided Adaptive Tree Rollout (PATR), a quality-aware rollout framework for multi-turn agent RL. PATR uses task-appropriate process feedback to score partial trajectories, selectively branches from promising states, reuses shared prefixes, and conservatively stops degenerate paths to reduce wasted sampling. The resulting rollout groups remain compatible with standard policy optimization while providing more efficient exploration under the same training budget. We evaluate PATR on FrozenLake and the challenging SWE-Bench, which is largely unexplored by prior tree-rollout agent RL methods. Experiments show that PATR improves performance by up to +5.0 points on SWE-Bench and +9.3 points on FrozenLake, highlighting process-guided tree rollouts as an effective strategy for scalable multi-turn RL.
Chinese Translation
强化学习（RL）已成为训练大规模语言模型（LLM）代理的关键方法，然而，像GRPO/RLOO这样的流行方法依赖于多条独立采样的完整轨迹进行优势估计。在长时间跨度的代理任务中，这种均匀的展开策略可能会在无信息的死胡同尝试上浪费预算，而有前景的中间状态却未能得到充分探索。代理轨迹的多回合结构，交错的动作和观察，自然支持将轨迹组组织为树形结构，其中每个回合作为探索的决策点。这一视角将有效探索重新框定为决定何处分支的问题。我们提出了过程评分引导的自适应树展开（Process-Scorer Guided Adaptive Tree Rollout, PATR），这是一个针对多回合代理强化学习的质量感知展开框架。PATR使用适合任务的过程反馈对部分轨迹进行评分，从有前景的状态选择性分支，重用共享前缀，并保守地停止退化路径以减少无效采样。最终生成的展开组与标准策略优化兼容，同时在相同的训练预算下提供更高效的探索。我们在FrozenLake和具有挑战性的SWE-Bench上评估PATR，后者在之前的树展开代理强化学习方法中几乎未被探索。实验表明，PATR在SWE-Bench上提高了最多5.0分，在FrozenLake上提高了最多9.3分，突显了过程引导的树展开作为可扩展多回合强化学习的有效策略。
cs.CL / 7 / 2607.15648
On the Structure of Address in Multi-Party Dialogue: From Discrete Labels to Continuous Levels
多方对话中地址结构的研究：从离散标签到连续层次
Abstract
In multi-party dialogues between a dialogue system and multiple users, identifying to whom an utterance is addressed is a key challenge. Prior work has typically treated addressee detection as a multi-class classification task, selecting a single label representing an individual participant or the group. This formulation assumes that address is inherently discrete and has primarily been used for predicting turn-taking. In this paper, we revisit this assumption by analyzing address as a continuous phenomenon. Using a multi-party human dialogue corpus annotated by multiple annotators, we construct both binary address labels derived from majority-vote addressee labels and continuous address levels inferred from annotator judgments using a latent-variable model. We then examine how these representations relate to turn-taking as well as listener behaviors, including gaze and backchannels. Our results show that, in addition to turn-taking, both gaze and backchannels are associated with address. Furthermore, models using continuous address levels achieve better predictive fit than those using discrete labels, suggesting that address may exhibit graded structure. Finally, we discuss the future directions of addressee detection research based on the findings of this study.
Chinese Translation
在对话系统与多个用户之间的多方对话中，识别发言对象是一个关键挑战。以往的研究通常将发言对象检测视为一个多类分类任务，选择一个代表单个参与者或群体的标签。这种表述假设地址本质上是离散的，并主要用于预测轮流发言。在本文中，我们通过将地址分析为一种连续现象来重新审视这一假设。我们使用一个由多个标注者注释的多方人类对话语料库，构建了基于多数投票的发言对象标签衍生的二元地址标签和基于标注者判断使用潜变量模型推断的连续地址层次。然后，我们考察这些表示与轮流发言以及听众行为（包括注视和反馈信号）之间的关系。我们的结果表明，除了轮流发言，注视和反馈信号也与地址相关。此外，使用连续地址层次的模型在预测拟合上优于使用离散标签的模型，暗示地址可能表现出分级结构。最后，我们基于本研究的发现讨论了发言对象检测研究的未来方向。
cs.CL / 8 / 2607.15655
Adaptive Multi-Step Lookahead Decoding for Diffusion Language Models
扩展多步前瞻解码用于扩散语言模型
Abstract
Masked diffusion language models (DLMs) enable parallel text generation by iteratively refining masked tokens, offering a promising alternative to autoregressive decoding. Recent lookahead-based decoding methods improve the accuracy--efficiency trade-off by exploring future decoding states before committing token updates. However, existing approaches mainly rely on shallow one-step lookahead, which optimizes immediate information gain but can be suboptimal for longer-horizon decoding trajectories. Meanwhile, we find that a naive extension for deeper lookahead is also ineffective, as fixed-depth rollout introduces additional computation and cannot adapt to heterogeneous intermediate decoding states. Thus, in this work, we propose AdaLook, an adaptive lookahead framework for DLM decoding. AdaLook dynamically determines whether to continue rollout based on candidate-score variance and further enables branch expansion when intermediate rollout states require additional exploration. This design avoids unnecessary deep rollout while allowing the decoder to re-trigger lookahead from informative intermediate states. Experiments on various benchmarks and models demonstrate that AdaLook achieves a better accuracy--decoding steps trade-off than existing one-step lookahead decoding methods.
Chinese Translation
掩码扩散语言模型（DLMs）通过迭代优化掩码标记实现并行文本生成，为自回归解码提供了一种有前景的替代方案。近期基于前瞻的解码方法通过在执行标记更新之前探索未来解码状态，改善了准确性与效率之间的权衡。然而，现有方法主要依赖于浅层的一步前瞻，这种方法优化了即时信息增益，但对于较长时间范围的解码轨迹可能并不理想。同时，我们发现简单扩展为更深的前瞻也无效，因为固定深度的展开引入了额外的计算，并且无法适应异质的中间解码状态。因此，在本研究中，我们提出了AdaLook，一个用于DLM解码的自适应前瞻框架。AdaLook动态决定是否继续展开，基于候选分数的方差，并在中间展开状态需要额外探索时进一步启用分支扩展。这一设计避免了不必要的深度展开，同时允许解码器从信息丰富的中间状态重新触发前瞻。在各种基准和模型上的实验表明，AdaLook在准确性与解码步骤之间的权衡上优于现有的一步前瞻解码方法。
cs.CL / 9 / 2607.15736
Better Starts, Better Ends: Bootstrapped Iterative Self-Reasoning Distillation for Compressed Reasoning
更好的开始，更好的结束：用于压缩推理的自举迭代自我推理蒸馏
Abstract
Large reasoning models often solve problems through long chain-of-thought (CoT) traces, yet much of this computation is spent on redundant derivations, repeated self-verification, and detours that do not improve the final answer. Existing on-policy self-distillation methods reduce this cost by matching a student model to a concise copy of itself on prefixes sampled from the student's own rollouts. We show that this objective has an initialization bottleneck. Since supervision is applied only to visited prefixes, training from a verbose base model places the KL loss on contexts that are often noisy, redundant, or already off track. In such regions, a concise teacher can provide only local corrections, while the student continues to explore trajectories that an efficient reasoner should avoid. In this paper, we propose BIRD(Bootstrapped Iterative Self-Reasoning Distillation), a two-stage self-reasoning distillation method that improves the rollout distribution before on-policy training. BIRD first samples concise solutions from the base model under a brevity instruction, keeps only answer-correct traces, and performs a lightweight prompt-switch SFT step. The traces are generated with the brevity instruction but learned under the original task prompt, turning instruction-induced conciseness into a default reasoning behavior. Starting from this warm model, BIRD then applies on-policy reverse-KL distillation with a concise self-teacher, now on cleaner and more informative prefixes. Across Qwen3 series models, BIRD achieves a stronger accuracy-efficiency trade-off than prompting and cold-start on-policy distillation on MATH-500 and AIME benchmarks. On Qwen3-8B, it improves MATH-500 accuracy from 86.2% to 92.0% while reducing the average response length from 3,099 to 1,115 tokens. These results highlight prefix support as a central factor in efficient reasoning distillation.
Chinese Translation
大型推理模型通常通过长链思维（CoT）轨迹来解决问题，但其中许多计算花费在冗余推导、重复自我验证和不改善最终答案的绕道上。现有的在线自蒸馏方法通过将学生模型与自身在学生自己回滚中采样的前缀的简洁副本进行匹配，从而减少了这一成本。我们表明，这一目标存在初始化瓶颈。由于监督仅应用于访问过的前缀，从冗长的基础模型进行训练会将KL损失施加在通常嘈杂、冗余或已经偏离轨道的上下文中。在这些区域，简洁的教师只能提供局部修正，而学生则继续探索高效推理者应避免的轨迹。在本文中，我们提出了BIRD（自举迭代自我推理蒸馏），这是一种两阶段自我推理蒸馏方法，在在线训练之前改善回滚分布。BIRD首先在简洁指令下从基础模型中采样简洁解决方案，仅保留答案正确的轨迹，并执行轻量级的提示切换SFT步骤。这些轨迹是在简洁指令下生成的，但在原始任务提示下学习，将指令引导的简洁性转化为默认的推理行为。在这个热身模型的基础上，BIRD随后应用在线反向KL蒸馏，使用简洁的自我教师，现在在更干净和更具信息性的前缀上进行。在Qwen3系列模型中，BIRD在MATH-500和AIME基准测试中实现了比提示和冷启动在线蒸馏更强的准确性与效率的权衡。在Qwen3-8B上，它将MATH-500的准确率从86.2%提高到92.0%，同时将平均响应长度从3,099个标记减少到1,115个标记。这些结果突显了前缀支持作为高效推理蒸馏的核心因素。
cs.CL / 10 / 2607.15766
Before the Action: Benchmarking LLMs on Prospective Hypothesis Discovery
行动之前：大型语言模型在前瞻性假设发现中的基准测试
Abstract
Large language models (LLMs) excel at answering pre-specified questions, yet their ability to navigate the open-ended, pre-conclusion stage of discovery remains largely unmeasured. We introduce Prospective Hypothesis Discovery (PHD), which asks models to autonomously construct grounded, discriminative, and testable hypothesis spaces from inconclusive evidence, including anomalous observations and fragmented records, to guide subsequent investigation. To evaluate this capability, we introduce HypoArena, comprising HypoData, a benchmark of 988 cases across six scientific and analytical domains, and HypoEval, an evaluation framework for open-ended hypothesis sets. To construct HypoData at scale, we propose Retrospective Context Regression, a Forge--Audit pipeline that reconstructs pre-conclusion contexts from completed expert documents by removing explicit conclusions, target hypotheses, and retrospective causal attributions while preserving the factual substrate. Because PHD admits multiple valid outputs, HypoEval combines bidirectional pairwise judgments with Bradley--Terry--Davidson aggregation for ranking and six-dimensional rubric scoring for diagnosis. Experiments on 15 frontier LLMs reveal clear capability stratification and model-dependent effects of structured analytical skills, with gains for several lower-performing models on HypoArena but regressions for other systems, including a top-performing model. Compared with absolute rubric scoring, arena evaluation resolves finer-grained differences among models, with aggregated rankings showing strong agreement with human experts and an independent judge. Together, these results support treating PHD as a distinct target for evaluating how LLMs formulate investigative directions when final conclusions are withheld. Our code and data are publicly available at github.com/SKYLENAGE-AI/HypoArena and github.com/SKYLENAGE-AI/HypoArena.
Chinese Translation
大型语言模型（LLMs）在回答预设问题方面表现出色，但它们在探索开放式、未得出结论的发现阶段的能力仍然未得到充分测量。我们引入了前瞻性假设发现（Prospective Hypothesis Discovery, PHD），该方法要求模型自主构建基于不确定证据（包括异常观察和零散记录）的有根据、可区分且可测试的假设空间，以指导后续调查。为了评估这一能力，我们推出了HypoArena，其中包含HypoData，这是一个涵盖六个科学和分析领域的988个案例的基准，以及HypoEval，一个用于开放式假设集的评估框架。为了大规模构建HypoData，我们提出了回顾性上下文回归（Retrospective Context Regression），这是一个Forge--Audit管道，通过去除明确结论、目标假设和回顾性因果归因，同时保留事实基础，从已完成的专家文档中重建未得出结论的上下文。由于PHD允许多个有效输出，HypoEval结合了双向成对判断与Bradley--Terry--Davidson聚合进行排名，并采用六维评分标准进行诊断。在对15个前沿LLM的实验中，显示出明显的能力分层和模型依赖的结构化分析技能效果，其中一些表现较差的模型在HypoArena上有所提升，但其他系统（包括一个表现最佳的模型）却出现了退步。与绝对评分标准相比，竞技场评估能够解决模型之间更细致的差异，聚合排名与人类专家和独立评审者之间显示出强一致性。综合这些结果，支持将PHD视为评估LLM在最终结论被保留时如何制定调查方向的一个独特目标。我们的代码和数据可在github.com/SKYLENAGE-AI/HypoArena和github.com/SKYLENAGE-AI/HypoArena上公开获取。
cs.CL / 11 / 2607.15829
Cost-efficient generative AI summarization for scalable automated essay scoring in educational assessment
成本效益高的生成式人工智能摘要用于可扩展的自动化作文评分在教育评估中的应用
Abstract
Automated essay scoring (AES) enables scalable assessment and timely feedback but remains challenged by transformer input-length limitations, which can cause information loss when processing long essays. This study proposes a generative AI-assisted summarization framework to improve long-form essay representation while maintaining scoring reliability. Using the ASAP 2.0 dataset, we generate controlled-length summaries with three GPT-5 variants (GPT-5, GPT-5 mini, and GPT-5 nano) and use them as inputs for downstream AES models. To preserve original writing signals, handcrafted linguistic features extracted from full essays are integrated with summary representations to form a hybrid framework. The approach is evaluated in terms of scoring performance, summarization quality, and computational cost. Scoring reliability is measured using quadratic weighted kappa (QWK), while summary quality is assessed through lexical overlap, semantic similarity, information retention, and redundancy metrics. Results show that GPT-5 mini achieves the highest agreement with human ratings, whereas GPT-5 produces the strongest summarization quality. Summary quality decreases for higher-scoring essays, indicating that more complex writing is more difficult to compress without information loss. These findings reveal trade-offs among model capacity, summary fidelity, cost efficiency, and preservation of educational constructs. This study provides an initial controlled evaluation of GPT-based summarization for AES and identifies important baselines and ablation studies required for future generalization. Overall, generative AI summarization offers a promising approach for scalable writing assessment while requiring careful validation of information preservation and fairness.
Chinese Translation
自动化作文评分（AES）能够实现可扩展的评估和及时反馈，但仍面临变换器输入长度限制的挑战，这在处理长篇作文时可能导致信息丢失。本研究提出了一种生成式人工智能辅助的摘要框架，以改善长篇作文的表现，同时保持评分的可靠性。使用ASAP 2.0数据集，我们生成具有控制长度的摘要，使用三种GPT-5变体（GPT-5、GPT-5 mini和GPT-5 nano）作为下游AES模型的输入。为了保留原始写作信号，我们将从完整作文中提取的手工语言特征与摘要表示相结合，形成一个混合框架。该方法在评分性能、摘要质量和计算成本方面进行了评估。评分可靠性使用二次加权卡帕（QWK）进行测量，而摘要质量则通过词汇重叠、语义相似性、信息保留和冗余度指标进行评估。结果表明，GPT-5 mini与人类评分的一致性最高，而GPT-5则产生了最强的摘要质量。对于高评分的作文，摘要质量下降，表明更复杂的写作在不丢失信息的情况下更难以压缩。这些发现揭示了模型能力、摘要保真度、成本效益和教育构念保留之间的权衡。本研究提供了对基于GPT的摘要在AES中应用的初步控制评估，并确定了未来推广所需的重要基准和消融研究。总体而言，生成式人工智能摘要为可扩展的写作评估提供了一种有前景的方法，同时需要对信息保留和公平性进行仔细验证。
cs.CL / 12 / 2607.15847
CAMMAR: Culture-Aware Matryoshka for Metaphorical Arabic Representations
CAMMAR：文化意识的套娃模型用于隐喻阿拉伯语表示
Abstract
Metaphor in Arabic is a culturally grounded mechanism for constructing meaning, encoding cultural knowledge that shapes interpretation. Yet current Arabic language models typically collapse lexical, cultural, and metaphorical information into a single representational space, a phenomenon we term "semantic smearing". We introduce CAMMAR (Culture-Aware Matryoshka for Metaphorical Arabic Representations), a representation learning framework that organizes meaning into nested lexical, cultural, and metaphorical embedding subspaces through a staged semantic curriculum. The design implements compositional principles of Al-Jurjani's theory of nazum, modeling figurative meaning as compositionally grounded in prior semantic relations, and yields a training-free geometric measure of metaphoricity based on the distance between lexical and metaphorical representations. Evaluated on a new span-annotated Arabic metaphor set as word-matched figurative/literal pairs, the geometric readout detects metaphor well above chance when the inter-layer geometry is shaped by paired supervision (AUC up to 0.84; figurative outscores its literal counterpart for the same word in 82.6\% of pairs), but sits at chance under an unsupervised domain contrast alone, a clean separation between a legible-under-supervision regime and a non-emergent one. A controlled ablation shows that grounding the lexical layer in morphological roots gives a small but consistent gain, an effect absent from direct probing that reflects the layer's quality as a measurement anchor. We will release the datasets, cultural concept inventory, and code upon acceptance.
Chinese Translation
阿拉伯语中的隐喻是一种文化根植的意义构建机制，编码了塑造解读的文化知识。然而，目前的阿拉伯语言模型通常将词汇、文化和隐喻信息压缩到一个单一的表征空间中，这一现象我们称之为“语义模糊”。我们提出了CAMMAR（Culture-Aware Matryoshka for Metaphorical Arabic Representations），这是一个表征学习框架，通过分阶段的语义课程将意义组织成嵌套的词汇、文化和隐喻嵌入子空间。该设计实现了Al-Jurjani的nazum理论的组合原则，将比喻意义建模为基于先前语义关系的组合基础，并基于词汇和隐喻表示之间的距离产生一种无训练的几何隐喻性度量。在一个新的跨度标注的阿拉伯隐喻集上进行评估，该集作为词匹配的比喻/字面对，几何读出在配对监督下的层间几何形状时，隐喻的检测显著高于随机水平（AUC高达0.84；在82.6%的对中，比喻的得分超过其字面对应词），但在仅有无监督领域对比时则处于随机水平，显示出在可监督的情况下与不可涌现的情况下的清晰分离。控制消融实验表明，将词汇层基于形态根进行基础化会带来小但一致的增益，而直接探测中缺乏这一效应，反映了该层作为测量锚的质量。我们将在接受后发布数据集、文化概念清单和代码。
cs.CL / 13 / 2607.15856
Contextual Semantic Relevance Tracks fMRI BOLD Responses During Naturalistic Speech Comprehension
上下文语义相关性追踪自然语言理解中的fMRI BOLD反应
Abstract
Naturalistic language comprehension requires listeners to process both local probabilistic expectations and contextual semantic relations. Surprisal has been widely used to quantify local word unexpectedness, but evidence that it robustly predicts fMRI BOLD responses during continuous comprehension has been mixed. This study investigates whether contextual semantic relevance, defined as how strongly an incoming word relates to its recent semantic context, predicts BOLD responses during naturalistic speech comprehension. We analyzed two public fMRI datasets, the Alice dataset and the Moth dataset, treating them as complementary rather than identical replications. Transformed BOLD responses were modeled with generalized additive mixed models (GAMMs) and original continuous BOLD time series were tested with FIR/deconvolution analyses. In Alice, semantic relevance was significant across all 12 ROIs (region of interest), whereas surprisal was not significant after FDR correction. In Moth, semantic relevance showed consistent negative effects across all 30 ROIs, while surprisal showed no comparable pattern. These findings suggest that semantic relevance is a promising BOLD-sensitive metric of contextual semantic fit. More broadly, our findings support the view that slow hemodynamic responses during naturalistic speech comprehension may be especially sensitive to contextual semantic integration, whereas local probabilistic prediction error may be more difficult to detect reliably with fMRI. In this sense, semantic relevance extends computational models of language comprehension from prediction alone toward context-sensitive semantic integration.
Chinese Translation
自然语言理解要求听者处理局部概率预期和上下文语义关系。惊讶度（Surprisal）已被广泛用于量化局部词汇的意外性，但在连续理解过程中，惊讶度是否能稳健预测fMRI BOLD反应的证据并不一致。本研究探讨了上下文语义相关性（定义为一个即将到来的词与其最近语义上下文的相关程度）是否能够预测自然语言理解中的BOLD反应。我们分析了两个公共fMRI数据集，Alice数据集和Moth数据集，将它们视为互补而非相同的重复实验。我们使用广义加性混合模型（GAMMs）对转换后的BOLD反应进行建模，并对原始连续BOLD时间序列进行FIR/去卷积分析。在Alice数据集中，语义相关性在所有12个感兴趣区域（ROI）中均显著，而惊讶度在FDR校正后并不显著。在Moth数据集中，语义相关性在所有30个ROI中显示出一致的负效应，而惊讶度则未显示出可比的模式。这些发现表明，语义相关性是一个有前景的BOLD敏感指标，能够反映上下文语义的适配性。更广泛地说，我们的发现支持这样一种观点：在自然语言理解过程中，缓慢的血流动力学反应可能对上下文语义整合特别敏感，而局部概率预测误差可能更难以通过fMRI可靠检测。从这个意义上说，语义相关性将语言理解的计算模型从单纯的预测扩展到了上下文敏感的语义整合。
cs.CL / 14 / 2607.15861
Conditional Reliability of Toxicity Signals for Multilingual and Code-Mixed Abuse Detection
多语言和代码混合滥用检测的毒性信号条件可靠性
Abstract
Moderation systems increasingly rely on external toxicity tools, but those tools are unreliable under code-mixing, transliteration, slang, and language mismatch. We study the \emph{conditional reliability} of toxicity priors in Indian multilingual and code-mixed short text: English toxicity, Indic abuse, and rule-based severity cues can be useful evidence, but only in some linguistic and abuse-severity contexts. We propose ToxGate, a trust-fusion head that conditions each auxiliary signal on the encoder representation before adding it to the prediction state. Across three short-text abuse datasets, four transformer encoders, and five seeds per setting, ToxGate improves over matched plain encoders in 10 of 12 in-domain settings and 7 of 8 transfer settings. The largest and most interpretable gains occur in high-risk moderation slices, including explicit slurs, violent threats, and cross-dataset transfer. The broader lesson is that moderation systems should treat external toxicity tools and priors as conditional evidence rather than fixed features or ground truth, in focused ablations, source-specific gating gives the strongest results in transfer, severe-abuse slices, and high-risk triage.
Chinese Translation
审核系统越来越依赖外部毒性工具，但这些工具在代码混合、音译、俚语和语言不匹配的情况下不可靠。我们研究了印度多语言和代码混合短文本中毒性先验的 extit{条件可靠性}：英语毒性、印度语言滥用和基于规则的严重性提示可以作为有用的证据，但仅在某些语言和滥用严重性上下文中有效。我们提出了ToxGate，这是一种信任融合头，它在将每个辅助信号添加到预测状态之前，基于编码器表示进行条件处理。在三个短文本滥用数据集、四个变换器编码器和每种设置五个种子下，ToxGate在12个领域内设置中改善了10个，在8个转移设置中改善了7个。最大的、最具可解释性的增益发生在高风险审核片段中，包括明确的侮辱、暴力威胁和跨数据集转移。更广泛的教训是，审核系统应将外部毒性工具和先验视为条件证据，而不是固定特征或真实情况，在集中消融中，源特定的门控在转移、严重滥用片段和高风险分流中给出了最强的结果。
cs.CL / 15 / 2607.15865
An MLIR-Based Compilation Method for Large Language Models
基于 MLIR 的大语言模型编译方法
Abstract
Large Language Models (LLMs) have become the dominant workload on modern AI accelerators, yet deploying them on specialized hardware still faces two core challenges: how to import a trained model into a compiler-friendly intermediate representation, and how to efficiently schedule the autoregressive inference loop under limited on-chip memory. This paper presents an MLIR (Multi-Level Intermediate Representation) based compilation method for large language models, illustrated using two dialects of operators, TopOp and TpuOp. TopOp serves as a high-level graph dialect that is independent of both the source framework and the target chip, and is responsible for expressing model semantics; TpuOp serves as the target hardware dialect, carrying chip-related decisions such as quantization, layer groups, and memory layout. A model is first represented as TopOp, then lowered layer by layer to TpuOp, and finally a deployable binary is generated. In addition, each Transformer layer is split into three stages for static compilation: prefill, prefill_kv (prefill with historical key-value cache), and decode, so as to accommodate the different computational characteristics of prompt-parallel processing and per-token generation. The method has been implemented in the TPU-MLIR compiler{https://github.com/sophgo/tpu-mlir} and the LLM-TPU deployment project\footnote{https://github.com/sophgo/LLM-TPU}, supporting a variety of generative models including the Qwen, Llama, InternVL, and MiniCPM-V series, as well as multiple quantization and deployment forms such as GPTQ, AWQ, and AutoRound.
Chinese Translation
大语言模型（LLMs）已成为现代人工智能加速器上的主导工作负载，但在专用硬件上部署它们仍面临两个核心挑战：如何将训练好的模型导入编译器友好的中间表示，以及如何在有限的片上内存下高效调度自回归推理循环。本文提出了一种基于 MLIR（多级中间表示）的大语言模型编译方法，使用了两种操作符方言，TopOp 和 TpuOp。TopOp 作为一种高层图方言，独立于源框架和目标芯片，负责表达模型语义；而 TpuOp 则作为目标硬件方言，承载与芯片相关的决策，如量化、层组和内存布局。模型首先表示为 TopOp，然后逐层降级为 TpuOp，最终生成可部署的二进制文件。此外，每个 Transformer 层被拆分为三个阶段进行静态编译：prefill、prefill_kv（带有历史键值缓存的预填充）和解码，以适应提示并行处理和逐令牌生成的不同计算特性。该方法已在 TPU-MLIR 编译器中实现，并支持多种生成模型，包括 Qwen、Llama、InternVL 和 MiniCPM-V 系列，以及多种量化和部署形式，如 GPTQ、AWQ 和 AutoRound。
cs.CL / 16 / 2607.15870
How Much Human Label Variation Does Formal Semantic Structure Explain?: Group-Level Effects and Item-Level Ceilings in NLI
正式语义结构解释多少人类标签变异？：NLI中的群体级效应和项目级上限
Abstract
Human label variation in natural language inference is increasingly treated as signal rather than noise, but how much of it formal semantic structure explains has not been measured directly. We measure it on the 3,113 SNLI and MNLI items of ChaosNLI, using a rule-based operator and monotonicity tagger validated against MED (0.883 agreement at the edit site, 0.807 on the sentence-level summary our analyses consume), three preregistered analysis blocks, and full reporting of negative results. Three bounds emerge. First, a group-level boundary: hypotheses that are not purely upward monotone show reliably higher label entropy (Cliff's delta = -0.284), and rank-based tests defend the effect against operator-presence and length reductions, though a bounded-outcome sensitivity check weakens the regression form of the length defense. Second, an item-level ceiling: the same formal profiles explain only 3.3 to 3.6 percent of entropy variance and reach a median-split AUC of 0.606, too weak to identify high-disagreement items. Third, composition invariance: across the boundary, three high-powered preregistered contrasts on validated error shares and explanation-type shares (VariErr, LiTEx) all return null results. In this sample, formal semantic structure shifts how much annotators disagree by a small amount and does not detectably change what they disagree about. ChaosNLI-S/M consists of items selected for low original agreement, and every claim is conditioned on that scope. All analyses were preregistered in a version-controlled research log, whose audit trail, including one corrected interpretation rule, the paper discloses.
Chinese Translation
在自然语言推理中，人类标签变异越来越被视为信号而非噪声，但正式语义结构解释了多少变异尚未直接测量。我们在ChaosNLI的3,113个SNLI和MNLI项目上进行测量，使用基于规则的运算符和单调性标记器，并通过MED进行验证（在编辑站点的协议率为0.883，在我们的分析所需的句子级总结中为0.807），并进行了三个预注册的分析块，全面报告负结果。出现了三个界限。首先，群体级边界：非纯上升单调的假设显示出可靠的更高标签熵（Cliff's delta = -0.284），基于排名的测试在运算符存在和长度减少的情况下支持该效应，尽管有界结果的敏感性检查削弱了长度防御的回归形式。其次，项目级上限：相同的正式特征仅解释了3.3%到3.6%的熵方差，并达到中位数分割AUC为0.606，太弱以识别高争议项目。第三，组合不变性：在该边界上，三个高效能的预注册对比在验证的错误份额和解释类型份额（VariErr, LiTEx）上均返回无效结果。在这个样本中，正式语义结构仅在小程度上改变了注释者之间的分歧，并未可检测地改变他们争论的内容。ChaosNLI-S/M由选择的低原始一致性项目组成，所有声明都以该范围为条件。所有分析均在版本控制的研究日志中预注册，该日志的审计跟踪，包括一个修正的解释规则，本文进行了披露。
cs.CL / 17 / 2607.15879
DECODEM: Data Extraction from Corporate Organizational Documents via Enhanced Methods
DECODEM：通过增强方法从企业组织文件中提取数据
Abstract
Much empirical legal research depends on translating unstructured text into structured variables. In corporate governance research as elsewhere, this translation has traditionally relied on human coding of documents such as charters and bylaws, a process that is costly, difficult to scale, and often opaque. This paper introduces DECODEM, a set of benchmark datasets for evaluating the automated extraction of corporate governance variables from organizational documents. The benchmarks pair randomly sampled corporate charters and bylaws with high-quality human annotations covering a range of governance provisions commonly studied in empirical work. Using these datasets, the paper evaluates several large-language-model extraction pipelines that vary in prompt design, task decomposition, and document handling. The underlying task consists of a set of document-level binary classification problems, one for each governance variable. The results show that automated extraction is feasible at a high level of accuracy for many provisions, with median performance near the upper bound across approaches. At the same time, performance varies systematically across variables, with a small number of provisions accounting for most of the remaining errors. More elaborate prompting strategies and cascading pipelines do not consistently improve performance for frontier models, but substantially narrow the gap between frontier and efficiency-oriented models in some settings, suggesting that pipeline design can partly substitute for model capability. By providing a standardized benchmark and a systematic evaluation of extraction methods, the paper demonstrates that current frontier models can extract legally meaningful information from complex corporate documents with high accuracy and suggests an important future role for automated feature extraction in constructing corporate governance datasets.
Chinese Translation
许多实证法律研究依赖于将非结构化文本转换为结构化变量。在公司治理研究及其他领域，这种转换传统上依赖于对章程和细则等文件的人工编码，这一过程成本高昂、难以扩展且往往不透明。本文介绍了DECODEM，一组用于评估从组织文件中自动提取公司治理变量的基准数据集。这些基准将随机抽样的公司章程和细则与涵盖实证研究中常见的治理条款的高质量人工注释配对。利用这些数据集，本文评估了几种大型语言模型提取管道，这些管道在提示设计、任务分解和文档处理上有所不同。基础任务由一组文档级二元分类问题组成，每个治理变量对应一个问题。结果表明，对于许多条款，自动提取在高准确率下是可行的，性能中位数接近各方法的上限。同时，性能在变量之间系统性地变化，少数条款占据了大部分剩余错误。更复杂的提示策略和级联管道并未始终提高前沿模型的性能，但在某些情况下显著缩小了前沿模型与效率导向模型之间的差距，这表明管道设计在一定程度上可以替代模型能力。通过提供标准化基准和对提取方法的系统评估，本文展示了当前前沿模型能够以高准确率从复杂的企业文件中提取法律上有意义的信息，并建议自动特征提取在构建公司治理数据集中的重要未来角色。
cs.CL / 18 / 2607.15893
Induction in Both Directions: A Mechanistic Analysis of In-Context Learning in Masked Diffusion Language Models
双向归纳：掩蔽扩散语言模型中上下文学习的机制分析
Abstract
While the internal mechanisms of autoregressive (AR) transformers have been studied extensively, much less is known about diffusion language models (DLMs), an emerging alternative that generates text by iterative denoising. In this work, we study how DLMs implement induction, a mechanism behind in-context learning in which the model finds a repeated context and copies the token that followed it. Our analysis compares attention-only AR models and absorbing-mask DLMs with matched architectures. We find that DLMs learn a bidirectional induction circuit, where previous-token and next-token heads write local context into the residual stream and later induction heads use it to find and copy the answer from the matching source position. The circuit is direction-symmetric, working whether the source appears in the past or in the future. When only left context is visible, matching what an AR model sees, the DLM does not outperform its AR counterpart in induction capabilities. However, we observe it has stronger induction when both sides of the masked token are visible, pointing to bidirectional context access rather than a stronger one-sided mechanism. Beyond induction, we provide causal evidence that DLMs compute the global fraction of masked tokens and use it as an implicit timestep, even though they are given no explicit timestep embedding.
Chinese Translation
尽管自回归（AR）变换器的内部机制已被广泛研究，但对于扩散语言模型（DLMs）这一新兴替代方案的了解却相对较少，该模型通过迭代去噪生成文本。在本研究中，我们探讨了DLMs如何实现归纳，这是一种在上下文学习中起作用的机制，其中模型找到重复的上下文并复制其后面的标记。我们的分析比较了仅关注的AR模型和具有匹配架构的吸收掩码DLMs。我们发现，DLMs学习了一个双向归纳电路，其中前标记和后标记头将局部上下文写入残差流中，随后归纳头利用这些信息从匹配源位置找到并复制答案。该电路是方向对称的，无论源出现在过去还是未来均有效。当仅可见左侧上下文时，匹配AR模型的视野，DLM在归纳能力上并未优于其AR对应物。然而，我们观察到当掩蔽标记的两侧均可见时，其归纳能力更强，这表明其具备双向上下文访问能力，而非单向机制的增强。除了归纳外，我们还提供了因果证据，表明DLMs计算被掩蔽标记的全局比例，并将其用作隐式时间步，即使它们没有被提供显式的时间步嵌入。
cs.CL / 19 / 2607.15957
From Plausible to Actionable: A Position on LLM Self-Explanations
从合理到可操作：关于大型语言模型自我解释的立场
Abstract
Large Language Models (LLMs) can generate natural language explanations that rationalize their own decisions, a phenomenon commonly referred to as self-explanations.Such explanations have emerged as a promising direction for explainable artificial intelligence (XAI), particularly for interpreting LLM behavior.However, while self-explanations often appear plausible, whether they faithfully reflect a model's underlying reasoning process remains an open question. In this opinion paper, we argue that self-explanations can be highly plausible, questionably faithful, and yet highly actionable. From a traditional XAI perspective, we identify the limitations of standard evaluation protocols for LLM-generated self-explanations and propose practical guidelines for assessing their plausibility and faithfulness. Moreover, we argue that evaluation should extend beyond these criteria to actionability, highlighting applications of LLM rationalization capabilities that support informed decision-making and appropriate action across diverse stakeholders.
Chinese Translation
大型语言模型（LLMs）能够生成自然语言解释，以合理化其自身的决策，这一现象通常被称为自我解释。这种解释已成为可解释人工智能（XAI）中一个有前景的方向，特别是在解释LLM行为方面。然而，尽管自我解释往往看起来合理，但它们是否忠实反映模型的潜在推理过程仍然是一个悬而未决的问题。在本文中，我们认为自我解释可以非常合理、可疑地忠实，但仍然具有高度的可操作性。从传统的XAI视角出发，我们识别出标准评估协议在评估LLM生成的自我解释方面的局限性，并提出评估其合理性和忠实性的实用指南。此外，我们认为评估应超越这些标准，关注可操作性，强调LLM合理化能力的应用，这些应用支持各利益相关者的知情决策和适当行动。
cs.CL / 20 / 2607.16001
BayesPO: Bayesian Prompt Optimization via Parallel-Tempered Gradient-Guided Discrete MCMC
BayesPO：通过并行温度梯度引导的离散MCMC进行贝叶斯提示优化
Abstract
Prompt optimization adapts large language models (LLMs) without updating model parameters, but many automatic prompt optimizers remain heuristic search procedures over candidate instructions. This paper studies prompt optimization as Bayesian posterior sampling over discrete prompt tokens. We define a posterior distribution by combining a task likelihood term, which rewards prompts that explain input-output examples, with a language-model prior, which favors fluent instructions. This converts prompt optimization into an energy-based posterior sampling problem, for which gradients can be used to guide discrete Markov chain Monte Carlo (MCMC) proposals over vocabulary tokens. We refer to our framework as BayesPO, short for Bayesian Prompt Optimization. In this paper, BayesPO is instantiated with Markov chain Monte Carlo: it uses a Metropolis-Hastings corrected Gibbs-with-Langevin (GwL) proposal and integrates parallel tempering for global exploration of rugged LLM-induced energy landscapes. The concrete sampler further adapts the GwL sampler to the practical constraints of non-weight-tied LLM embeddings. Experiments with Qwen2.5 models show that the sampler discovers semantically meaningful prompts on diagnostic tasks, that parallel tempering helps escape a local optimum in a poetry completion task, and that post-optimizing APE prompts on 24 instruction-induction subtasks improves average accuracy from 60.04% to 63.23%. The study also reveals two main limitations: energy minimization may overfit small optimization sets, and the current sampler remains computationally expensive. These findings position Bayesian prompt sampling as a principled post-optimization tool and point to a promising direction for probabilistic prompt optimization.
Chinese Translation
提示优化在不更新模型参数的情况下调整大型语言模型（LLMs），但许多自动提示优化器仍然是对候选指令的启发式搜索过程。本文将提示优化视为对离散提示标记的贝叶斯后验采样。我们通过结合任务似然项（奖励能够解释输入输出示例的提示）和语言模型先验（偏好流畅指令）来定义后验分布。这将提示优化转化为一个基于能量的后验采样问题，其中梯度可以用于指导对词汇标记的离散马尔可夫链蒙特卡洛（MCMC）提议。我们将我们的框架称为BayesPO，即贝叶斯提示优化的缩写。在本文中，BayesPO通过马尔可夫链蒙特卡洛实现：它使用经过Metropolis-Hastings修正的Gibbs-with-Langevin（GwL）提议，并结合并行温度进行对崎岖的LLM引导的能量景观的全局探索。具体的采样器进一步将GwL采样器适应于非权重绑定LLM嵌入的实际约束。与Qwen2.5模型的实验表明，采样器在诊断任务中发现了语义上有意义的提示，并且并行温度有助于在诗歌完成任务中逃离局部最优解，此外，在24个指令诱导子任务上对APE提示进行后优化将平均准确率从60.04%提高到63.23%。研究还揭示了两个主要限制：能量最小化可能会对小的优化集过拟合，且当前的采样器仍然计算开销较大。这些发现将贝叶斯提示采样定位为一个有原则的后优化工具，并指向概率提示优化的一个有前景的方向。
cs.CL / 21 / 2607.16021
Candidate Attended Dialogue State Tracking Using BERT
基于BERT的候选人参与对话状态跟踪
Abstract
Dialogue state tracking (DST) is one of the core components in task-oriented dialogue systems. At each turn in a conversation, DST estimates the user belief or dialogue state, which is used as input for downstream modules to predict system actions and generate responses. The increasingly popular dialogue system applications like Google Assistant, Siri and Alexa need to support a large number of services and APIs, resulting in growing attention to the scalability of such systems. Especially for some domains with little or no training data, the capability of transferring existing knowledge of other domains is highly desired. In this paper, we present a novel scalable framework for multi-domain dialogue state tracking. The proposed system leverages the pretrained BERT model to achieve zero-shot generalization, making it easy to quickly adapt to new domains without additional training. The performance of our model is evaluated on recently released schema-based dialogue (SGD) dataset, showing significant improvement compared to previous baseline.
Chinese Translation
对话状态跟踪（DST）是任务导向对话系统的核心组成部分。在对话的每个回合中，DST 估计用户的信念或对话状态，这些状态作为下游模块的输入，用于预测系统动作和生成响应。随着 Google Assistant、Siri 和 Alexa 等对话系统应用的日益普及，这些系统需要支持大量的服务和 API，从而引起了对其可扩展性的关注。尤其是在一些训练数据稀缺或没有的领域，转移其他领域现有知识的能力是非常受欢迎的。本文提出了一种新颖的多领域对话状态跟踪可扩展框架。所提出的系统利用预训练的 BERT 模型实现零样本泛化，使其能够快速适应新领域而无需额外训练。我们在最近发布的基于模式的对话（SGD）数据集上评估了模型的性能，结果显示与之前的基线相比有显著改善。
cs.CL / 22 / 2607.16051
Loop the Loopies!
环形循环模型！
Abstract
We present Loopie, the most powerful looped Transformer to date. The Loopie series consists of two Mixture-of-Experts (MoE) models: a 20B-parameter model with 2B active parameters and a 6Bparameter model with 0.6B active parameters. Looped Transformers have long faced a challenge: given an N-fold increase in pre-training compute, increasing the parameter count by a factor of N usually outperforms looping a model N times. Loopie addresses this challenge. Extensive ablation studies, including comparisons with a vanilla 30B-A3B model, show that Loopie substantially outperforms vanilla Transformer baselines trained with the same compute budget. Our novel post-training pipeline equips Loopie with strong reasoning abilities. At the 2025 IMO and IPhO, Loopie achieves gold-medal performance without tools.
Chinese Translation
我们提出了Loopie，这是迄今为止最强大的循环变换器（looped Transformer）。Loopie系列包含两个专家混合模型（Mixture-of-Experts, MoE）：一个拥有200亿参数的模型，其中20亿参数为活跃参数；另一个是一个拥有60亿参数的模型，其中6亿参数为活跃参数。循环变换器长期以来面临一个挑战：在预训练计算量增加N倍的情况下，将参数数量增加N倍通常比将模型循环N次的效果更好。Loopie解决了这一挑战。广泛的消融研究，包括与一个基础的30B-A3B模型的比较，表明Loopie在相同计算预算下显著优于基础变换器基线。我们新颖的后训练流程使Loopie具备了强大的推理能力。在2025年国际数学奥林匹克（IMO）和国际物理奥林匹克（IPhO）中，Loopie在没有工具的情况下实现了金牌表现。
cs.CL / 23 / 2607.16057
Frontier AI performance across the business disciplines: a case-grounded benchmark of knowledge work and analytical reasoning
前沿人工智能在商业学科中的表现：基于案例的知识工作和分析推理基准
Abstract
Large language models (LLMs) are improving rapidly as reflected in benchmark scores, yet these AI benchmarks largely test capabilities such as factual recall, narrow question answering, mathematical problem-solving, and coding and agentic tool-use. What remains poorly measured is AI progress on the analytical knowledge work white-collar professionals perform daily, including synthesizing complex information, exercising judgment under uncertainty and incomplete information, applying strategic and adversarial thinking in multi-stakeholder settings, weighing trade-offs, and producing defensible, structured analyses. This gap is even more pronounced for subjective components of such work, where success can be challenging to define. The "case method" form of education practiced by top business schools provides a natural foundation for addressing this measurement gap, and we construct BusinessCaseBench, a benchmark spanning hundreds of questions drawn from business cases across eighteen disciplines, each paired with a grading rubric derived from the expert-written instructor case solution. On BusinessCaseBench, frontier AI models already score highly against instructor rubrics, and capability within one model family improves substantially over two years. These results provide strong evidence that AI performance on this class of work is already high and rapidly improving, with implications for business schools, where case pedagogy trains undergraduates and MBAs in this kind of analytical reasoning, and for entry-level professional roles, where such skills have historically anchored early-career work.
Chinese Translation
大型语言模型（LLMs）正在快速提升，这在基准得分中得到了体现。然而，这些人工智能基准主要测试诸如事实回忆、狭义问答、数学问题解决以及编程和工具使用等能力。相对而言，人工智能在白领专业人士日常进行的分析性知识工作方面的进展仍然测量不足，包括综合复杂信息、在不确定和不完整信息下做出判断、在多方利益相关者环境中应用战略和对抗性思维、权衡取舍以及产生可辩护的结构化分析。这一差距在此类工作的主观成分上尤为明显，因为成功的定义往往具有挑战性。顶级商学院所实践的“案例教学法”为解决这一测量差距提供了自然的基础，我们构建了BusinessCaseBench，这是一个涵盖来自十八个学科的商业案例的数百个问题的基准，每个问题都配有来自专家撰写的教师案例解决方案的评分标准。在BusinessCaseBench上，前沿人工智能模型在教师评分标准下已经取得了高分，并且同一模型家族的能力在两年内显著提升。这些结果强有力地表明，人工智能在这一类工作的表现已经很高并且正在快速改善，这对商学院具有重要意义，因为案例教学法培养本科生和MBA学生进行这种类型的分析推理，同时也对入门级专业角色产生影响，因为这些技能历来是早期职业工作的基础。
cs.CL / 24 / 2607.16072
Frontier Language Models Struggle to Copy: Text Can Be Better Viewed in 2D
前沿语言模型在复制方面的困难：文本可以更好地以二维方式呈现
Abstract
While large language models (LLMs) can solve advanced reasoning problems in seconds, we show that even frontier models fail to perform a much simpler operation: exactly copying an input string that lies well within their context windows. We attribute this failure to positional encodings in Transformer architectures, whose inductive bias favors copying through a shortcut based on matching local contexts rather than carefully locating the corresponding input positions. To address this issue, we introduce 2D-RoPE, which organizes text into a 2D grid rather than a 1D sequence and assigns each token a row ID and a column ID. Under this view, copying becomes simply retrieving input tokens at a fixed column offset, which makes the task easy to learn. In synthetic copy experiments, shallow Transformers with 2D-RoPE achieve perfect copying at input lengths hundreds of times longer than those seen during training, whereas standard positional encodings fall far behind. We further show that the advantage of 2D-RoPE language models on copy tasks consistently holds in large-scale pretraining on DCLM with model sizes up to 1.4B parameters. Overall, our results suggest that viewing text in 2D can benefit language modeling, and we hope this encourages future work to further explore the potential of 2D positional encodings.
Chinese Translation
尽管大型语言模型（LLMs）能够在几秒钟内解决复杂的推理问题，但我们展示了即使是前沿模型也无法执行一个更简单的操作：精确复制一个位于其上下文窗口内的输入字符串。我们将这一失败归因于Transformer架构中的位置编码，其归纳偏差倾向于通过基于匹配局部上下文的捷径进行复制，而不是仔细定位相应的输入位置。为了解决这个问题，我们引入了2D-RoPE，它将文本组织成一个二维网格，而不是一维序列，并为每个标记分配一个行ID和一个列ID。在这种视角下，复制变成了简单地在固定列偏移量处检索输入标记，这使得任务易于学习。在合成复制实验中，使用2D-RoPE的浅层Transformer在输入长度远远超过训练期间所见的情况下实现了完美复制，而标准位置编码则远远落后。我们进一步展示了在DCLM上进行大规模预训练时，2D-RoPE语言模型在复制任务上的优势在模型规模达到14亿参数时仍然保持一致。总体而言，我们的结果表明，以二维方式查看文本可以对语言建模带来益处，我们希望这能鼓励未来的研究进一步探索二维位置编码的潜力。
cs.CL / 25 / 2607.16085
Controlling Implicit Shortcut Reliance in L2 Spoken English Auto-markers
控制第二语言口语英语自动评分中的隐性捷径依赖
Abstract
Increasingly, speech and language processing tasks take either audio or text directly rather than extracting features from these as the input to the classifier or regressor. Often these systems make use of complex, for example transformer-based, processes that have the ability to derive highly non-linear mappings between the input and the output. Unfortunately these systems can also learn ''shortcuts'' where the classifier is overly reliant on particular aspects of the input to yield the output. For the task of language proficiency assessment, this over-reliance can enable learners to increase their score by exploiting the shortcut rather than improving their ability. This paper introduces a novel training criterion that is able to reduce the classifier's reliance on shortcuts, thus for example limiting this option for malpractice in language assessment. This process is illustrated on two forms of assessment system, one based on the audio the other on the speech recognition text. The results show that, for both systems, there is higher correlations with features that could be exploited for malpractice than expected from the human reference, indicating an over-reliance on these features. By introducing the modified training criterion, this correlation can be reduced to be closer to the reference correlation.
Chinese Translation
越来越多的语音和语言处理任务直接使用音频或文本，而不是从中提取特征作为分类器或回归器的输入。这些系统通常利用复杂的过程，例如基于变换器（transformer）的过程，能够推导输入与输出之间高度非线性的映射。不幸的是，这些系统也可能学习到“捷径”，使得分类器过于依赖输入的特定方面来产生输出。在语言能力评估任务中，这种过度依赖可能使学习者通过利用捷径来提高分数，而不是提升自身能力。本文提出了一种新颖的训练标准，能够减少分类器对捷径的依赖，从而限制在语言评估中的不当行为选项。该过程在两种评估系统上进行了说明，一种基于音频，另一种基于语音识别文本。结果表明，对于这两种系统，能够被不当利用的特征与人类参考之间的相关性高于预期，表明对这些特征的过度依赖。通过引入修改后的训练标准，这种相关性可以减少到更接近参考相关性的水平。
cs.CL / 26 / 2607.16117
Rate-Utility Frontiers for Language Encodings: Comparing Tokens, Bytes, and Pixels Under Controlled Linguistic Content
语言编码的速率-效用边界：在受控语言内容下比较标记、字节和像素
Abstract
Language models encode text as subword tokens, raw bytes, or rendered pixels, but these encodings are usually compared under modeling constraints that expose different amounts of linguistic content to models across different languages. We instead ask what each encoding preserves when both the content and the downstream capacity are controlled. Using verified parallel sentences across thirteen languages and five scripts, we compare tokens, bytes, and pixels through a shared bottleneck whose width is swept to trace rate-utility frontiers. This separates three quantities that are often conflated: the number of input positions an encoding creates, the latent capacity available after encoding, and the task-relevant information that survives compression. We evaluate three utilities: surface form preservation, cross-lingual sentence alignment, and topic classification. No encoding dominates across tasks or capacity regimes. Pixels preserve surface form best, bytes preserve cross-lingual alignment best, especially in same-script multilingual settings, and tokens support topic prediction best. These performances are not explained by sequence length alone. Short inputs can discard useful meaning, while long inputs can preserve information that compresses well. Choosing an encoding is therefore not a fixed preference for tokens, bytes, or pixels, but a rate-utility tradeoff that depends on the task, language mix, capacity regime, and compute budget.
Chinese Translation
语言模型将文本编码为子词标记、原始字节或渲染像素，但这些编码通常在建模约束下进行比较，这些约束在不同语言中向模型暴露了不同量的语言内容。我们则探讨在控制内容和下游能力的情况下，每种编码所保留的内容。通过使用跨十三种语言和五种书写系统的经过验证的平行句子，我们通过一个共享的瓶颈比较标记、字节和像素，其宽度被调整以描绘速率-效用边界。这一过程区分了三个常常混淆的量：编码所产生的输入位置数量、编码后可用的潜在容量，以及在压缩中存活的与任务相关的信息。我们评估了三种效用：表面形式保留、跨语言句子对齐和主题分类。在任务或容量模式中，没有一种编码在所有方面都占优。像素在表面形式保留方面表现最佳，字节在跨语言对齐方面表现最佳，尤其是在同书写系统的多语言环境中，而标记在主题预测方面表现最佳。这些表现并非仅由序列长度解释。短输入可能会丢失有用的含义，而长输入则可能保留压缩效果良好的信息。因此，选择编码并不是对标记、字节或像素的固定偏好，而是一个依赖于任务、语言组合、容量模式和计算预算的速率-效用权衡。
cs.CL / 27 / 2607.16131
ToolSciVer: Multimodal Scientific Claim Verification with Visual Tool Augmented Reinforcement Learning
ToolSciVer：基于视觉工具增强的多模态科学声明验证
Abstract
Multimodal Scientific Claim Verification (MSCV) requires models to verify scientific claims using visually grounded evidence from papers, including figures, tables, charts, and textual context. However, existing methods often fail because they struggle to locate decisive visual evidence, accurately read structured scientific visuals, and integrate multimodal observations into reliable reasoning. We introduce ToolSciVer, the first tool-augmented framework for MSCV to our knowledge. ToolSciVer equips a VLM with three type-aware visual tools, table row/column focus, chart-to-structure parsing, and high-resolution region zoom, which convert dense scientific visuals into explicit, claim-facing evidence, and trains the policy with Group Relative Policy Optimization (GRPO) under a composite reward of answer correctness, format validity, length control, tool-use efficiency, and tool-validity penalties. Experiments on SciVer and MuSciClaims datasets on five VLMs from three model families (Qwen, InternVL, Gemma) demonstrate that our method achieves superior performance compared to four competitive baselines including prompting-based and RL-based tool-use methods, highlighting the effectiveness of learned, type-aware tool use for scientific claim verification.
Chinese Translation
多模态科学声明验证（MSCV）要求模型使用来自论文的视觉证据（包括图形、表格、图表和文本上下文）来验证科学声明。然而，现有方法往往失败，因为它们难以定位决定性的视觉证据，准确读取结构化的科学视觉信息，并将多模态观察整合为可靠的推理。我们介绍了ToolSciVer，这是我们所知的第一个用于MSCV的工具增强框架。ToolSciVer为视觉语言模型（VLM）配备了三种类型感知的视觉工具：表格行/列聚焦、图表到结构解析和高分辨率区域放大，这些工具将密集的科学视觉信息转换为明确的、面向声明的证据，并通过群体相对策略优化（GRPO）在答案正确性、格式有效性、长度控制、工具使用效率和工具有效性惩罚的复合奖励下训练策略。在SciVer和MuSciClaims数据集上的实验，使用来自三个模型系列（Qwen、InternVL、Gemma）的五个VLM，证明了我们的方法在性能上优于包括基于提示和基于强化学习的工具使用方法在内的四个竞争基线，突显了学习的、类型感知的工具使用在科学声明验证中的有效性。