Skip to content

[AAAI 2023] The official implementation of "A Benchmark and Asymmetrical-Similarity Learning for Practical Image Copy Detection"

License

Notifications You must be signed in to change notification settings

WangWenhao0716/ASL

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

17 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

A Benchmark and Asymmetrical-Similarity Learning for Practical Image Copy Detection (AAAI 2023)

The proposed NDEC dataset

image

Training set

The 100,000 negative pairs

The original 100,000 training images from ISC2021: Please refer to Meta AI download page

Reference set

The original 100,000 reference images from ISC2021: Please refer to Meta AI download page

Query set

The collected 49,252 query images

Groudtruth

The groundtruth file for the query set

ASL

image

Step 1

Please first review the code of two baselines in our paper from D^2LV and BoT.

Step 2

Please refer to the train.py for the norm-ratio based loss.

Step 3

Please integrate the train.py into the training code of Step 1.

Step 4

Please refer to the test.py to find our how we perform testing.

Step 5

Please integrate the test.py into the test code of Step 1.

Step 6 (Optional)

You can also integrate train.py and test.py to other baselines, such as EsViTp, CNNCL, and EfNet.

Citation

If you use the proposed dataset or find the ASL helpful, please cite our work:

@inproceedings{wang2023benchmark,
  title={A benchmark and asymmetrical-similarity learning for practical image copy detection},
  author={Wang, Wenhao and Sun, Yifan and Yang, Yi},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={37},
  number={3},
  pages={2672--2679},
  year={2023}
}

Our work is mainly based on our two former competition (ISC2021) winner solutions:

@article{wang2021d,
  title={D\^{} 2LV: A Data-Driven and Local-Verification Approach for Image Copy Detection},
  author={Wang, Wenhao and Sun, Yifan and Zhang, Weipu and Yang, Yi},
  journal={arXiv preprint arXiv:2111.07090},
  year={2021}
}
@article{wang2021bag,
  title={Bag of Tricks and A Strong baseline for Image Copy Detection},
  author={Wang, Wenhao and Zhang, Weipu and Sun, Yifan and Yang, Yi},
  journal={arXiv preprint arXiv:2111.08004},
  year={2021}
}

The ISC2021 competition conclusion:


@InProceedings{pmlr-v176-papakipos22a,
  title = 	 {Results and findings of the 2021 Image Similarity Challenge},
  author =       {Papakipos, Zo\"e and Tolias, Giorgos and Jenicek, Tomas and Pizzi, Ed and Yokoo, Shuhei and Wang, Wenhao and Sun, Yifan and Zhang, Weipu and Yang, Yi and Addicam, Sanjay and Papadakis, Sergio Manuel and Ferrer, Cristian Canton and Chum, Ond{\v{r}}ej and Douze, Matthijs},
  booktitle = 	 {Proceedings of the NeurIPS 2021 Competitions and Demonstrations Track},
  pages = 	 {1--12},
  year = 	 {2022},
  editor = 	 {Kiela, Douwe and Ciccone, Marco and Caputo, Barbara},
  volume = 	 {176},
  series = 	 {Proceedings of Machine Learning Research},
  month = 	 {06--14 Dec},
  publisher =    {PMLR},
  pdf = 	 {https://proceedings.mlr.press/v176/papakipos22a/papakipos22a.pdf},
  url = 	 {https://proceedings.mlr.press/v176/papakipos22a.html},
}

About

[AAAI 2023] The official implementation of "A Benchmark and Asymmetrical-Similarity Learning for Practical Image Copy Detection"

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages