PhysCaP：以物理信息探索錨定程式碼即策略代理

摘要

我們提出 PhysCaP，一種用於機器人操作中主動感知的具物理資訊之程式碼即策略（Code-as-Policy）代理。雖然視覺-語言-動作策略擅長模仿示範，但它們依賴被動觀察，無法推斷對操作至關重要的潛在物理屬性。PhysCaP 在程式碼即策略框架中加入了物理資訊探索層，使代理能透過互動進行明確的資訊搜尋。它引入了免訓練的物理屬性提取模組，僅憑機器人本體感覺即可估算物體質量與剛度，無需額外感測器。為了平衡探索成本與所獲資訊的效率，PhysCaP 採用雙代理設計：規劃器（Planner）決定何時探索、何時停止；優先排序器（Prioritizer）過濾不合理的互動，並以啟發式優先分數對其餘互動進行排序，從而實現高效且有目標的探索。我們在真實世界的桌面操作任務（搜尋隱藏物體、偵測空罐、尋找成熟酪梨）以及 LIBERO 中的模擬任務上評估 PhysCaP。結果顯示，現有的被動式與樸素互動式基準方法在物理屬性隱藏時不是失敗，就是過度探索；而 PhysCaP 能以較少的互動次數和縮短的執行時間達到相當的效能。消融研究進一步驗證了所提出之物理屬性提取模組的有效性。專案頁面：https://physcap.github.io

English

We present PhysCaP, a Physics-Informed Code-as-Policy agent 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. It 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 dual-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 real-world tabletop manipulation tasks (searching for hidden objects, detecting empty cans, and finding ripe avocados) 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