Zixuan Chen^*,1 | Xialin He^*,2 | Yen-Jen Wang^*,3 | Qiayuan Liao³ | Yanjie Ze⁴ | Zhongyu Li³
S. Shankar Sastry³ | Jiajun Wu⁴ | Koushil Sreenath³ | Saurabh Gupta² | Xue Bin Peng^1,5
1 Simon Fraser University   ² UIUC   ³ UC Berkeley   ⁴ Stanford University   ⁵ NVIDIA
* Equal Contribution

This codebase supports simulation training of five different humanoid robots: Fourier GR1T1, Fourier GR1T2, Unitree H1, Berkeley Humanoid, and Unitree G1. The simulation training is based on Isaac Gym. We also provide a sim-to-sim pipeline for these robots. The sim-to-sim is performed in Mujoco. Please find the detailed instructions for simulation training in simulation.

Our codebase currently provides the real-world deployment code of Fourier GR1T2. Please find the detailed instructions for real-world deployment in deployment.

• We would like to thank all the authors in this project, this project cannot be finished without your efforts!
• Our simulation environment implementation is based on legged_gym, and the rl algorithm implementation is based on rsl_rl.
• Humanoid-Gym and Expressive-Humanoid also provide lots of insights.

If you find this work useful, please consider citing:

@article{chen2024lcp,
title={Learning Smooth Humanoid Locomotion through Lipschitz-Constrained Policies},
author={Zixuan Chen and Xialin He and Yen-Jen Wang and Qiayuan Liao and Yanjie Ze and Zhongyu Li and S. Shankar Sastry and Jiajun Wu and Koushil Sreenath and Saurabh Gupta and Xue Bin Peng},
journal={arxiv Preprint arXiv:2410.11825},
year={2024}
}