Awesome
<h1 align="center"> Open-Vocabulary DETR with Conditional Matching </h1> <h2 align="center"> <a href="https://arxiv.org/pdf/2203.11876.pdf">arXiv</a> | <a href="https://www.mmlab-ntu.com/project/ovdetr/index.html">Project Page</a> | <a href="https://github.com/yuhangzang/OV-DETR">Code</a> </h2>This repository contains the implementation of the following paper:
<p align="center"> <img width=95% src="./assets/framework.png"> </p>Open-Vocabulary DETR with Conditional Matching<br> Yuhang Zang, Wei Li, Kaiyang Zhou, Chen Huang, Chen Change Loy<br> European Conference on Computer Vision (ECCV), 2022<br>
Installation
We use the same environment as Deformable DETR. You are also required to install the following packages:
We test our models under python=3.8, pytorch=1.11.0, cuda=10.1, 8 Nvidia V100 32GB GPUs.
Data
Please refer to dataset_prepare.md.
Running the Model
Please refer to run_scripts.md.
Model Zoo
- Open-vocabulary COCO (AP50 metric)
| Base | Novel | All | Model |
|---|---|---|---|
| 61.0 | 29.4 | 52.7 | Google Drive |
Citation
If you find our work useful for your research, please consider citing the paper:
@InProceedings{zang2022open,
author = {Zang, Yuhang and Li, Wei and Zhou, Kaiyang and Huang, Chen and Loy, Chen Change},
title = {Open-Vocabulary DETR with Conditional Matching},
journal = {European Conference on Computer Vision},
year = {2022}
}
License
<a rel="license" href="http://creativecommons.org/licenses/by-nc-sa/4.0/"><img alt="Creative Commons License" style="border-width:0" src="https://i.creativecommons.org/l/by-nc-sa/4.0/88x31.png" /></a><br />This work is licensed under a <a rel="license" href="http://creativecommons.org/licenses/by-nc-sa/4.0/">Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License</a>.
Acknowledgement
We would like to thanks Deformable DETR, CLIP and ViLD for their open-source projects.
Contact
Please contact Yuhang Zang if you have any questions.