arXiv:2611.05754 · v1

GRASP: Gradient-Guided Sampling for Contact-Rich Physics-Based Motion Retargeting

首次提交 2026-11-04 07:53

版本更新 2026-11-04 07:53

雷达观测 2026-11-05 10:08

摘要

We study physics-based retargeting of human demonstrations for robot learning: finding robot controls that reproduce reference motions under contact dynamics. We propose GRASP (Gradient Recovery from Archived Samples for Proposals), which combines sampling-based exploration with gradient-guided refinement without requiring a differentiable simulator. GRASP-TO extends tempered sequential Monte Carlo by fitting local quadratic cost models to archived rollouts and using their gradients to guide Markov chain Monte Carlo rejuvenation. Metropolis-Hastings correction preserves the target distribution for each fixed proposal kernel, and a population analysis connects the recovered gradients to Gaussian smoothing. GRASP-MPC scales this optimizer to long demonstrations through warm-started receding-horizon planning. Experiments on dexterous hands and humanoids show accurate retargeting with a shared tracking objective and no task-specific contact rewards, including 92.8% success on 1352 bimanual OakInk2 demonstrations. The refined references improve PPO tracking in pen spinning and humanoid-object interaction. We also demonstrate dexterous manipulation and whole-body humanoid motion on real hardware.

作者备注：26 pages, including references and appendices. Project website: https://computationalrobotics.seas.harvard.edu/GRASP/