Guanqi He(@guanqi_he):TopoRetarget enables a dexterous manipulation RL mimic pipeline from human demonstrations to real-world robot execution. With TopoRetarget + RL tracking, we achieve zero-shot sim-to-real pen spinning on the Wuji Hand.🎉🎉🎉 We collect human hand-object motion with Wuji Glove, retarget the human demonstration to wuji hand while preserving hand-object interaction information, and train a RL policy to track the resulting hand-object reference. Paper: https://t.co/e9TrpkFMpd Project page:

Guanqi He

@guanqi_he

PhD at Tsinghua Univeristy, Wuji Tech @wuji_global | Prev. Amazon FAR, CMU @LeCARLab, @AirLabCMU | control and robotics

参加 June 2023

255 フォロー中 528 ファン

Guanqi He@guanqi_he

2026.06.17 08:52

TopoRetarget enables a dexterous manipulation RL mimic pipeline from human demonstrations to real-world robot execution. With TopoRetarget + RL tracking, we achieve zero-shot sim-to-real pen spinning on the Wuji Hand.🎉🎉🎉 We collect human hand-object motion with Wuji Glove, retarget the human demonstration to wuji hand while preserving hand-object interaction information, and train a RL policy to track the resulting hand-object reference. Paper: Project page: