AI RESEARCH
PhyGile: Physics-Prefix Guided Motion Generation for Agile General Humanoid Motion Tracking
arXiv CS.AI
•
ArXi:2603.19305v1 Announce Type: cross Humanoid robots are expected to execute agile and expressive whole-body motions in real-world settings. Existing text-to-motion generation models are predominantly trained on captured human motion datasets, whose priors assume human biomechanics, actuation, mass distribution, and contact strategies. When such motions are directly retargeted to humanoid robots, the resulting trajectories may satisfy geometric constraints (e.g., joint limits and pose continuity) and appear kinematically reasonable.