DPText-DETR (mirror)

This repository is an archival mirror of pretrained weights from the official DPText-DETR repository. No changes have been made to the weights themselves; they are re-hosted here for long-term preservation as part of a personal project.

• Original paper: Ye, M., Zhang, J., Zhao, S., Liu, J., Du, B., & Tao, D. (2023). DPText-DETR: Towards Better Scene Text Detection with Dynamic Points in Transformer. AAAI 2023 (Oral).
• Original code: https://github.com/ymy-k/DPText-DETR
• Checkpoint in this repo: art_final.pth - the fine-tuned model for the ICDAR19 ArT benchmark, as released by the original authors.
• All credit for the model architecture, training, and weights belongs to the original authors. This repo makes no claim of authorship over the model.

License

These weights are distributed under the original AdelaiDet non-commercial license (see the LICENSE file in this repo). In short:

• Free to use, copy, and redistribute for non-commercial / research purposes, provided the copyright notice and disclaimer are retained.
• Commercial use is not permitted under this license. For a commercial license, contact the original AdelaiDet authors (chhshen@gmail.com), as noted in the LICENSE file.

This mirror does not alter or override those terms.

Citation

If you use this model, please cite the original paper:

@inproceedings{ye2022dptext,
  title={DPText-DETR: Towards Better Scene Text Detection with Dynamic Points in Transformer},
  author={Ye, Maoyuan and Zhang, Jing and Zhao, Shanshan and Liu, Juhua and Du, Bo and Tao, Dacheng},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={37},
  number={3},
  pages={3241--3249},
  year={2023}
}