Beyond Kinematic Tracking: Preserving Expressive Characteristics in Humanoid Whole-Body Control

This research introduces a residual learning framework with compensation objectives targeting amplitude, energy, body expansion, body lean, and pelvis height to preserve expressive characteristics in humanoid whole-body control. Motion-aware gating is used to allocate correction capacity based on body-part activity, and motion-specific residual policies are distilled into a unified policy. Simulation and physical experiments on the Unitree G1 across eight motion categories demonstrate superior preservation of dynamic and postural motion characteristics compared to baselines.