ResearchBenchmarksRL Motion Imitation on Unitree H1 Balancing Stick 2Follow364.5Final RewardDDR248.124278.337308.55338.763May 22, 2026Evaluation ResultsMethodMethodLinksFinal RewardSteps to 90% RewardDDRReference Type=Direct...Reference Type=Direct Dynamic Retargeting2026.05364.578,900GRReference Type=Kinemat...Reference Type=Kinematic Retargeting2026.05252.6103,200