WARP: Whole-Body Retargeting for Learning from Offline Human Demonstrations

2026-06-29

AI Summary

This work addresses the embodiment discrepancy problem in transferring offline human demonstrations to full-body robot motion, where existing retargeting methods often produce multimodal and inconsistent motions. The authors propose WARP, an offline pipeline that, for the first time, enables zero-shot, human-in-the-loop-free full-body mobile manipulation. By explicitly modeling morphological differences between humans and robots, integrating a closed-form shoulder-elbow-wrist (SEW) geometric solver to ensure precise end-effector tracking, and introducing inertial mobile base control to maintain whole-body motion consistency, WARP directly generates unique and accurate robot trajectories from offline demonstrations. This approach supports high-fidelity open-loop execution on real robots and provides high-quality data for supervised policy learning, significantly outperforming current retargeting methods.

Abstract

Direct transfer from human demonstration to learnable robot action is a crucial step towards scalable whole-body mobile manipulation. While human data scales better than mobile teleoperation, it requires overcoming significant embodiment gaps. Existing retargeting methods yield imprecise or inconsistent solutions, causing action multi-modality that prevents supervised policies from reliably converging. We present Whole-body-Aware Retargeting from human Pose (WARP), an offline pipeline that explicitly models embodiment differences to extract precise, unique whole-body actions. WARP leverages a closed-form Shoulder-Elbow-Wrist (SEW) geometric solver for exact end-effector tracking while preserving whole-body structural intent. Paired with lazy mobile-base control, it extracts accurate, consistent robot trajectories. Evaluations show WARP provides highly reliable data for open-loop real-world replay. To our knowledge, WARP is the first framework to achieve zero-shot whole-body mobile manipulation directly from offline human demonstrations, eliminating the need for human-in-the-loop teleoperation action data. More details on https://warp-retarget.github.io/

Problem

• embodiment gap
• whole-body retargeting
• offline human demonstrations
• action multi-modality
• mobile manipulation

Innovation

• whole-body retargeting
• embodiment gap
• geometric solver
• offline imitation learning
• zero-shot manipulation