YangCaoCS/codav2_released_models

The pretrained models of CoDAv2 on SUN-RGBD and ScanNet.

If they are helpful, please cite:

@inproceedings{cao2023coda,
  title={CoDA: Collaborative Novel Box Discovery and Cross-modal Alignment for Open-vocabulary 3D Object Detection},
  author={Cao, Yang and Zeng, Yihan and Xu, Hang and Xu, Dan},
  booktitle={NeurIPS},
  year={2023}
}

@article{cao2024collaborative,
  title={Collaborative Novel Object Discovery and Box-Guided Cross-Modal Alignment for Open-Vocabulary 3D Object Detection},
  author={Cao, Yang and Zeng, Yihan and Xu, Hang and Xu, Dan},
  journal={IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)},
  year={2025}
}