From Festival Video to the Stage: A Sim-to-Real Ladder for Long-Horizon Expressive Whole-Body Dance on a Commercial Humanoid
From Festival Video to the Stage: A Sim-to-Real Ladder for Long-Horizon Expressive Whole-Body Dance on a Commercial Humanoid
This research presents a pipeline that transfers the traditional Japanese Awa Odori festival dance from monocular video to a commercial Unitree G1 EDU humanoid robot.
The study demonstrates a pipeline transferring traditional Japanese Awa Odori dance from monocular video to a Unitree G1 EDU humanoid using reinforcement learning and kinematic retargeting, achieving multiple rope-free performance durations including a continuous 9.7-minute routine.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
Reported numbers
performance duration
43
View original evidence
ent ladder evaluated on hardware examines three interventions. We detect and repair silent corruption of deployment artifacts, model state-estimator drift during training, and edit the reference motion to support stable execution and a controlled ending. The final policies achieved three consecutive rope-free completions of a 43-second routine, a 46-second performance with automatic handoff, and a continuous 9.7-minu
Open source S3performance duration
46
View original evidence
ent ladder evaluated on hardware examines three interventions. We detect and repair silent corruption of deployment artifacts, model state-estimator drift during training, and edit the reference motion to support stable execution and a controlled ending. The final policies achieved three consecutive rope-free completions of a 43-second routine, a 46-second performance with automatic handoff, and a continuous 9.7-minu
Open source S3performance duration
9.7
View original evidence
ent ladder evaluated on hardware examines three interventions. We detect and repair silent corruption of deployment artifacts, model state-estimator drift during training, and edit the reference motion to support stable execution and a controlled ending. The final policies achieved three consecutive rope-free completions of a 43-second routine, a 46-second performance with automatic handoff, and a continuous 9.7-minu
Open source S3torque-free window
45–140
View original evidence
te rope-free performance composed from the captured motion. In a live handover, a single remote request let the robot start from and return to the manufacturer's walking controller on stage, with a measured torque-free window of 45-140 ms. We report the training dynamics and hardware outcomes and discuss failure boundaries involving compliant flooring and occlusion in monocular capture.
Open source S4
- paper: not_reported
- code: not_reported
- weights: not_reported
- dataset: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
ent ladder evaluated on hardware examines three interventions. We detect and repair silent corruption of deployment artifacts, model state-estimator drift during training, and edit the reference motion to support stable execution and a controlled ending. The final policies achieved three consecutive rope-free completions of a 43-second routine, a 46-second performance with automatic handoff, and a continuous 9.7-minu
Open source S3
te rope-free performance composed from the captured motion. In a live handover, a single remote request let the robot start from and return to the manufacturer's walking controller on stage, with a measured torque-free window of 45-140 ms. We report the training dynamics and hardware outcomes and discuss failure boundaries involving compliant flooring and occlusion in monocular capture.
Open source S4
Source：arXiv Robotics — research abstracts · arxiv.org