发表机构
National Taiwan University; NVIDIA Research; National Yang Ming Chiao Tung University（台湾大学; 英伟达研究院; 国立阳明交通大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
PhysCaP是一种带物理引导探索层的代码即策略智能体，采用双智能体设计，可高效估算物体物理属性，在桌面操纵及LIBERO模拟任务中性能优于基线方法。
AI 中文摘要
我们提出了PhysCaP，一种用于机器人操纵主动感知的物理引导代码即策略（Code-as-Policy）智能体。尽管视觉-语言-动作策略在模仿演示方面表现出色，但它们依赖被动观察，无法推断操纵所需的潜在物理属性。PhysCaP在代码即策略框架中增加了物理引导探索层，可通过交互实现显式信息获取。它引入了无需训练的物理属性提取模块，该模块仅通过机器人本体感知即可估算物体质量和刚度，无需额外传感器。为平衡探索成本与获取信息的效率，PhysCaP采用双智能体设计：规划器（Planner）决定何时探索、何时停止；优先级器（Prioritizer）过滤不合理交互，并使用启发式优先级分数对剩余交互排序，从而实现高效、定向的探索。我们在真实桌面操纵任务（搜索隐藏物体、检测空罐子、寻找成熟牛油果）及LIBERO模拟任务中对PhysCaP进行评估。结果显示，现有被动及朴素交互基线要么在物理属性隐藏时失效，要么过度探索，而PhysCaP以更少的交互次数和更短的执行时间取得了相当的性能。消融研究进一步验证了所提物理属性提取模块的有效性。项目页面：this https URL
英文摘要
We present PhysCaP, a Physics-Informed Code-as-Policy agent system for active perception in robotic manipulation. While vision-language-action policies excel at imitating demonstrations, they rely on passive observation and fail to infer latent physical properties critical for manipulation. PhysCaP augments code-as-policy frameworks with a physics-informed exploration layer that enables explicit information-seeking through interaction. Our method introduces training-free physical property extraction modules that estimate object mass and stiffness from robot proprioception without additional sensors. To balance exploration costs and the efficiency of information obtained, PhysCaP employs a multi-agent design: a Planner that decides when to explore and when to stop, and a Prioritizer that filters implausible interactions and ranks the remainder using a heuristic priority score, enabling efficient, targeted exploration. We evaluate PhysCaP on three real-world tabletop manipulation tasks and a simulated task in LIBERO. The results show that existing passive and naive interactive baselines either fail when physical properties are hidden or over-explore, whereas PhysCaP achieves comparable performance with fewer interactions and reduced execution time. Ablation studies further validate the effectiveness of the proposed physical property extraction modules. Project page: https://physcap.github.io
