Adapt Human Motion for Robots with NVIDIA SOMA Retargeter
Web page · https://developer.nvidia.com/blog/adapt-human-motion-for-robots-with-nvidia-soma-retargeter
Human motion offers roboticists a great starting point for teaching humanoid robots how to move. A human demonstration captures the coordination and timing...
First seen by Oriel on .
AI summary
NVIDIA SOMA Retargeter is an open-source tool that converts SOMA human-motion BVH animations into robot-specific joint trajectories, supporting configuration, refinement, batch processing, and downstream motion-policy training.
- It scales targets to robot proportions, solves inverse-kinematics objectives using Newton and NVIDIA Warp, applies post-processing, and exports root poses and joint angles as CSV.
- SOMA Retargeter v0.2 includes configurations and reference results for Unitree G1, Unitree H2, Booster T1, AgiBot X2 Ultra, and AgiBot A3T3.
- Robot Configurator supports reference-pose alignment, skeleton-to-body mapping, reusable configurations, foot-plant correction, and experimental IK weight optimization.
- Retargeted trajectories feed SONIC or ProtoMotions training workflows, with simulation used to evaluate balance, contacts, actuator limits, and tracking accuracy.
#motion retargeting #humanoid robots #inverse kinematics #robot learning
Generated by gpt-5.6-sol. Check the original for details.
Aliases
| Source | External ID | URL |
|---|---|---|
| NVIDIA Developer Blog | https://developer.nvidia.com/blog/?p=125128 | https://developer.nvidia.com/blog/adapt-human-motion-for-robots-with-nvidia-soma-retargeter/ Read discussion |