The Invisible EgoHand: 3D Hand Forecasting through EgoBody Pose Estimation

This is the official code releasse for "The Invisible EgoHand: 3D Hand Forecasting through EgoBody Pose Estimation".

🔨 Installation

git clone https://github.com/masashi-hatano/EgoH4.git
cd EgoH4
python3 -m venv egoh4
source egoh4/bin/activate
pip install -r requirements.txt

🔥 Training

Please run the following command to train the EgoH4 model. The checkpoint can be downloaded from here.

python3 lit_main.py train=True test=False

🔍 Evaluation

To evaluate the model, please run the following command.

python3 lit_main.py train=False test=True devices=[0] strategy=auto

✍️ Citation

If you find our work useful in your research, please consider citing:

@article{Hatano2025EgoH4,
  title={The Invisible EgoHand: 3D Hand Forecasting through EgoBody Pose Estimation},
  author = {Hatano, Masashi and Zhu, Zhifan and Saito, Hideo and Damen, Dima},
  journal={arXiv preprint arXiv:2504.08654},
  year={2025}
}