← 研踪 PaperTrailarXiv:2611.22713 · v1
Body Schema Policy: Body as the Vision-Action Interface for Efficient Robot Loco-Manipulation
首次提交 2026-11-17 20:52
版本更新 2026-11-17 20:52
雷达观测 2026-11-18 11:33
摘要
Currently, robots can learn whole-body loco-manipulation through scale: the strongest policies have billions of parameters and still need task-specific fine-tuning. Humans have limited processing capacity, yet learn and act efficiently through a body schema that organizes what they see and how they move. We introduce Body Schema Policy (BSP), an imitation learning policy that builds such a schema into the vision-action interface. Its latent body schema places joint and link tokens on the robot's kinematic tree, where learned screw motions carry features between links. Through body-queried vision, each body part retrieves its own visual evidence, which conditions the denoising of its actions. With this inductive bias, BSP returns an action chunk faster than any policy we compare. Compact, body-specific visual readouts also emerge without spatial supervision: manipulation commands depend strongly on task objects, while locomotion commands respond more broadly to the scene. On nine simulated Unitree G1 whole-body tasks, BSP matches the humanoid foundation model with fewer parameters and nearly doubles Diffusion Policy's success. The same holds on a real G1 and a Fetch mobile manipulator.
观测到的版本
相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。
v1 · 2026-11-17 20:52 · 最新观测版本
Body Schema Policy: Body as the Vision-Action Interface for Efficient Robot Loco-Manipulation
Currently, robots can learn whole-body loco-manipulation through scale: the strongest policies have billions of parameters and still need task-specific fine-tuning. Humans have limited processing capacity, yet learn and act efficiently through a body schema that organizes what they see and how they move. We introduce Body Schema Policy (BSP), an imitation learning policy that builds such a schema into the vision-action interface. Its latent body schema places joint and link tokens on the robot's kinematic tree, where learned screw motions carry features between links. Through body-queried vision, each body part retrieves its own visual evidence, which conditions the denoising of its actions. With this inductive bias, BSP returns an action chunk faster than any policy we compare. Compact, body-specific visual readouts also emerge without spatial supervision: manipulation commands depend strongly on task objects, while locomotion commands respond more broadly to the scene. On nine simulated Unitree G1 whole-body tasks, BSP matches the humanoid foundation model with fewer parameters and nearly doubles Diffusion Policy's success. The same holds on a real G1 and a Fetch mobile manipulator.
查看此版本解读 →查看此版本原文 ↗