git clone https://github.com/lmy1001/One2Any.git
cd One2Any
conda env create --file env.yaml
conda activate one2any_env

For model training, you need both oo3d_9d_dataset and foundationpose_dataset.

For oo3d_9d_dataset, please follow here for data download and preparation.

For foundationpose_dataset, please follow here for data download and preparation.

./train.sh

Here is an example of evaluation on linemod dataset. You have to first download dataset from BOP benchmark.

The pretrained model can be downloaded here, put in under ./pretrained_model/, and run

./test.sh

If you find this project useful in your research, please cite:

@inproceedings{liu2025one2any,
  title={One2Any: One-Reference 6D Pose Estimation for Any Object},
  author={Liu, Mengya and Li, Siyuan and Chhatkuli, Ajad and Truong, Prune and Van Gool, Luc and Tombari, Federico},
  booktitle={Proceedings of the Computer Vision and Pattern Recognition Conference},
  pages={6457--6467},
  year={2025}
}

This project is developed upon OV9D, Oryon. We thank the authors for open sourcing their great works!