CoDance: Learning Reactive and Compliant Human-Humanoid Interaction from Video

2026-10-04

AI Summary

This study addresses the challenge of coordinating humanoid robot motion and responding to interaction forces from human dance partners during sustained physical contact. To this end, it proposes a force-aware imitation learning framework based on video-based motion retargeting and multi-link compliance augmentation. This approach transforms kinematic demonstrations into training data incorporating structured external forces, combined with reinforcement learning to train reactive, compliant co-dancing policies. In simulation, the method reproduces approximately 80% of wrist displacement. On physical hardware, it achieves sustained bimanual co-dancing with humans and adaptively transitions between forward and backward movements. These results effectively resolve the force-motion coordination problem inherent in human-robot physical interaction.

Abstract

Partnered human-humanoid interaction couples locomotion with continuous physical contact. A humanoid needs to coordinate with a person's motion while responding to interaction forces and maintaining stable and natural movement. We present CoDance, a framework for learning reactive and compliant human-humanoid interaction from video. We study partnered dancing as a challenging instantiation, where a humanoid coordinates its footsteps with a moving partner and maintains continuous two-hand contact. Given a single video of two human dancers, CoDance retargets their motions into a robot reference and a moving partner. We introduce a multi-link compliance augmentation that transforms the kinematic demonstration into force-aware training data by adapting the robot reference under structured forces at both hands. Policies trained on this data follow the observed partner while preserving the demonstrated locomotion style and responding compliantly to physical interaction. In simulation, the policies adapt their footsteps to changes in the partner and reproduce approximately 80% of the wrist displacement encoded by the augmented demonstrations. On a physical humanoid, CoDance enables sustained two-hand dancing with a human partner including repeated transitions between forward and backward motions.

Problem

• Human-Humanoid Interaction
• Partnered Dancing
• Physical Contact
• Compliant Motion
• Locomotion Coordination

Innovation

• Human-Humanoid Interaction
• Multi-link Compliance Augmentation
• Motion Retargeting
• Force-aware Policy Learning
• Partnered Dancing