Temporal Morphological-Symmetry-Equivariant Heterogeneous Graph Neural Network for Robotic Dynamics Learning
This repository is an extension of the L4DC 2025 paper: Morphological-Symmetry-Equivariant Heterogeneous Graph Neural Network for Robotic Dynamics Learning (Project Page | arXiv) as part of a university project. We extend the paper by encoding temporal relationships in two different ways: 1) space-time-graphs and 2) adding temporal cross- and self-attention.
To get started, setup a Conda Python environment with Python=3.11:
conda create -n ms-hgnn python=3.11
conda activate ms-hgnn
Then, install the library (and dependencies) with the following command:
cd MorphSym-HGNN
pip install .
Note, if you have any issues with setup, refer to environment_files/README.md so you can install the exact libraries we used.
The necessary URDF files are part of git submodules in this repository, so run the following commands to download them:
git submodule init
git submodule update