Installation Guide

Follow these steps to set up your environment for GenManip.

1. Install Dependencies

GenManip requires Conda for environment management and specific Python packages, including PyTorch and Isaac Sim.

1. Create and activate a Conda environment. We recommend Python 3.10.

   conda create -n genmanip python=3.10
   conda activate genmanip

2. Install PyTorch. Ensure you install the version compatible with your CUDA toolkit (cu121 shown here).

   pip install torch==2.4.0 --index-url https://download.pytorch.org/whl/cu121

3. Install Isaac Sim and related packages. Use the NVIDIA PyPI index.

   pip install isaacsim==4.1.0 isaacsim-extscache-physics==4.1.0 isaacsim-extscache-kit==4.1.0 isaacsim-extscache-kit-sdk==4.1.0 --extra-index-url https://pypi.nvidia.com

4. Install other required Python packages.

   pip install mplib open3d roboticstoolbox-python shapely trimesh python-fcl lmdb concave_hull

5. Clone the GenManip repository.

   git clone https://github.com/OpenRobotLab/GenManip
   cd genmanip

2. Download and Prepare Assets

You need to download the scene assets and layout files for the benchmarks.

1. Download Scene Assets. The assets contain the 3D models and textures used in the benchmark scenes.

   Asset	Description	Download Link
   GenManipBench-42	The first subset of GenManipBench, contains 42 scenes	TBD
   GenManipBench-200	The full version of GenManipBench, contains 200 scenes	TBD

   Unzip the downloaded file(s) to the genmanip_assets directory in the root of the repository.

2. Post-process Assets. Run the provided script to prepare the downloaded assets (USD files) for use with Isaac Sim. Replace /path/to/your/assets with the actual path where you unzipped the assets.

   # Process the base asset file
   python standalone_tools/usda_gen.py -f genmanip_assets/
   # Process the benchmark scene files
   python standalone_tools/usda_gen.py -f genmanip_assets/GenManipBench_final

   Note: Ensure you run the script on both the base.usd file and the directory containing the numbered scene USD files.

3. Download Layout Files. Layout files define the initial object poses and configurations for each benchmark task.

   Asset	Description	Download Link
   Layout-42	The layouts for the first subset of GenManipBench (42 scenes)	TBD
   Layout-200	The layouts for the full version of GenManipBench (200 scenes)	TBD

   Unzip the downloaded layout files to the saved/layouts directory in the root of the repository.

3. Verify Installation (Render Test)

To ensure your installation and environment are set up correctly, run the render_test.py script. This script will attempt to load a basic scene and render an image.

python render_test.py

If the script executes without errors, it will save a rendered image test.png to a tmp directory within your project folder. Successfully generating this image indicates that Isaac Sim is likely installed and configured correctly.

You are now ready to configure your environment and run the benchmarks. Proceed to the Quick Start guide.