Please email zongzhuofan@gmail.com to obtain the model weights and configs. When submitting your email, please provide the following details:
Name: (Kindly provide your full name.)
Affiliation: (Specify the name or URL of your university or company.)
Job Title: (For example, professor, PhD, researcher, etc.)
Email: (Your email address.)
Usage Intent: (For non-commercial purposes.)
The Co-DETR model on the COCO dataset will be released in the near future.
Model | Backbone | Aug | Dataset | box AP (val) | box AP (minival) | box AP (test-dev) |
---|---|---|---|---|---|---|
Co-DINO | ViT-L | DETR | COCO | 65.9 | - | 66.0 |
Co-DINO | ViT-L | LSJ | LVIS | 68.0 | 72.0 | - |
If you find this model useful, please use the following BibTeX entry for citation.
@misc{codetr2022,
title={DETRs with Collaborative Hybrid Assignments Training},
author={Zhuofan Zong and Guanglu Song and Yu Liu},
year={2022},
eprint={2211.12860},
archivePrefix={arXiv},
primaryClass={cs.CV}
}