Bi-manual robotic teleoperation system using VR hand tracking and camera streaming. Saves data to .h5
files for training. Visualizes the robot using rerun
.
git clone https://github.com/kscalelabs/arm-teleop.git && cd arm-teleop
conda create -y -n arm-teleop python=3.8 && conda activate arm-teleop
pip install -r requirements.txt
Start the server on the robot computer.
python play.py --robot="5dof"
Start ngrok on the robot computer.
ngrok http 8012
Open the browser app on the HMD and go to the ngrok URL.
Start the server on the robot computer.
🤗 K-Scale HuggingFace Datasets
- Vuer is used for visualization
- PyBullet is used for IK
- ngrok is used for networking
- Rerun is used for visualization
- H5Py is used for logging datasets
- HuggingFace is used for dataset & model storage
@misc{teleop-2024,
title={Bi-Manual Remote Robotic Teleoperation},
author={Hugo Ponte},
year={2024},
url={https://github.com/kscalelabs/teleop}
}