Garment Support#
In RoboVerse, we also integrate garment and deformable object manipulation tasks into the RoboVerse benchmark. We build upon the foundations of GarmentLab. While we initially attempted to migrate these tasks as-is from the main branch, we encountered an incompatibility issue with the PhysX simulation backend between OmniIsaacGymEnvs and the provided garment simulation. So, garment manipulation currently is supported on garment
branch. As the community’s research on soft body control in simulations progresses, we will continue to update this branch.
🛠️ Get Started#
Follow the steps below to get the Garment Support feature up and running:
1. Install Isaac Sim 4.0.0#
This branch is built with Isaac Sim 4.0.0, which is required due to a deprecation warning for particle cloth in Isaac Sim 4.1.0 (used in the main branch).
For convenience, you can add the following alias to your .bashrc
or .zshrc
to use the correct Python environment:
alias omni_python="~/.local/share/ov/pkg/isaac-sim-4.0.0/python.sh"
2. Download the Garment Assets#
Make sure to download the required GarmentLab assets and extract them into the Assets
folder within your project directory. Ensure that the folder structure matches the original asset directory and that all files are correctly placed.
3. Install the Additional Dependencies#
Once the assets are in place, you’ll need to install some additional dependencies. Run the following command:
omni_python -m pip install open3d
4. Launch the Demo Collection#
The demo scripts are located in the scripts
folder. You can launch the demo with the following command:
omni_python demo/HangDemo.py --usd_path $usd_path --visual_material_path $visual_material_path
Then, the rendered data will be stored in the same location as in the main branch.
💡 Notes:
usd_path
: The path to the USD file for the garment (e.g.,Assets/Garment/Dress/.../garment.usd
).visual_material_path
: The path to the material file for visual textures (e.g.,Assets/Material/.../material.usd
).
As we progress with garment manipulation research, we will continue updating this repository and resolving existing issues.
🛠️ Troubleshooting#
If you encounter any issues with the setup, please open an issue in the repository. We appreciate your feedback as we continue to refine and improve this integration!
Feel free to ask if you need further assistance or details! 😊