SLR: Learning Quadruped Locomotion without Privileged Information

Official Implementation for SLR: Learning Quadruped Locomotion without Privileged Information.

🌟 2024 Conference on Robot Learning

This codebase is licensed under the CC BY-NC 4.0 license, with inherited licenses from Legged Gym and RSL RL by ETH Zurich, Nikita Rudin and NVIDIA CORPORATION & AFFILIATES. Commercial use of the material is prohibited.

Please read through the entire README before cloning the repo.

---

🛠️ Getting Started

🚀 Pipeline to Install and Train SLR

Note: Before running our code, it's highly recommended to first familiarize yourself with the RSL's Legged Gym version to get a basic understanding of the Isaac-LeggedGym-RslRL framework.

---

1. Create an environment and install PyTorch

conda create -n slr python=3.8
conda activate slr
pip3 install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu116

---

2. Install Isaac Gym (Preview 4 release)

Download and install from Isaac Gym:

cd isaacgym/python && pip install -e .

Test the installation by running an example:

cd examples && python 1080_balls_of_solitude.py

---

3. Clone this repository and install packages:

git clone [email protected]:11chens/SLR-master.git

Ensure you have installed the following packages:

cd SLR-master && pip install .

---

4. Try Training the Model

You can use the --headless flag to disable the GUI. Press "v" to pause/resume GUI playback.

python train.py

---

5. Play the Trained Policy

python play.py

---

📑 Citation

If our work has helped you, please consider citing the following:

@inproceedings{
chen2024slr,
title={{SLR}: Learning Quadruped Locomotion without Privileged Information},
author={Shiyi Chen and Zeyu Wan and Shiyang Yan and Chun Zhang and Weiyi Zhang and Qiang Li and Debing Zhang and Fasih Ud Din Farrukh},
booktitle={8th Annual Conference on Robot Learning},
year={2024},
url={https://openreview.net/forum?id=RMkdcKK7jq}
}

---

👏 Acknowledgements

This project leverages components from several amazing open-source repositories:

• Legged Gym
• RSL RL
• Isaac Gym
• LocomotionWithNP3O
• ABS

Thanks to the authors and contributors for their work!