Name
stringlengths
9
9
Level
int64
0
5
generated_images
imagewidth (px)
512
512
real_images
imagewidth (px)
150
8.1k
CAP000008
2
CAP000013
4
CAP000018
0
CAP000025
1
CAP000030
1
CAP000031
2
CAP000035
3
CAP000051
2
CAP000065
1
CAP000066
1
CAP000074
0
CAP000076
3
CAP000089
2
CAP000095
0
CAP000113
3
CAP000115
2
CAP000136
2
CAP000156
2
CAP000162
0
CAP000164
3
CAP000179
0
CAP000225
5
CAP000258
4
CAP000264
0
CAP000282
0
CAP000294
3
CAP000303
1
CAP000305
2
CAP000307
4
CAP000329
5
CAP000331
2
CAP000337
1
CAP000340
5
CAP000341
1
CAP000347
4
CAP000353
5
CAP000357
3
CAP000364
4
CAP000390
0
CAP000396
4
CAP000397
3
CAP000401
3
CAP000424
3
CAP000436
5
CAP000440
5
CAP000455
5
CAP000477
3
CAP000481
1
CAP000493
4
CAP000513
2
CAP000522
1
CAP000544
2
CAP000550
2
CAP000557
1
CAP000564
5
CAP000568
2
CAP000569
2
CAP000595
0
CAP000598
0
CAP000628
3
CAP000635
0
CAP000642
0
CAP000645
1
CAP000652
4
CAP000653
4
CAP000669
2
CAP000682
3
CAP000684
3
CAP000686
1
CAP000689
4
CAP000699
1
CAP000707
3
CAP000713
5
CAP000714
5
CAP000722
0
CAP000727
4
CAP000734
2
CAP000744
1
CAP000789
3
CAP000803
1
CAP000810
2
CAP000812
3
CAP000819
4
CAP000824
0
CAP000831
0
CAP000842
0
CAP000853
3
CAP000865
0
CAP000885
0
CAP000890
3
CAP000891
4
CAP000892
5
CAP000899
2
CAP000905
0
CAP000927
0
CAP000932
5
CAP000934
1
CAP000935
1
CAP000939
3
CAP000949
4

Summary

This is the dataset proposed in our paper Image Copy Detection for Diffusion Models (NeurIPS 2024).

D-Rep consists of 40, 000 image-replica pairs, in which each replica is generated by a diffusion model. The 40, 000 image-replica pairs are manually labeled with 6 replication levels ranging from 0 (no replication) to 5 (total replication). We divide D-Rep into a training set with 90% (36, 000) pairs and a test set with the remaining 10% (4, 000) pairs.

Download

Automatical

Install the datasets library first, by:

pip install datasets

Then it can be downloaded automatically with

from datasets import load_dataset
dataset = load_dataset('WenhaoWang/D-Rep')

Manual

You can also download each file by wget:

wget https://huggingface.co/datasets/WenhaoWang/D-Rep/resolve/main/training_pairs.tar
wget https://huggingface.co/datasets/WenhaoWang/D-Rep/resolve/main/test_pairs.tar
wget https://huggingface.co/datasets/WenhaoWang/D-Rep/resolve/main/labels.csv

Curators

D-Rep is created by Wenhao Wang, Dr. Yifan Sun, Zhentao Tan and Professor Yi Yang.

License

We release our D-Rep under the CC-BY-NC-4.0 license.

Helpful Links

The project homepage: https://icdiff.github.io/

The code of image copy detection for diffusion models: https://github.com/WangWenhao0716/PDF-Embedding

The official reviews of our paper: https://openreview.net/forum?id=gvlOQC6oP1

The Arxiv: https://arxiv.org/abs/2409.19952

Citation

@article{wang2024icdiff,
  title={Image Copy Detection for Diffusion Models},
  author={Wang, Wenhao and Sun, Yifan and Tan, Zhentao and Yang, Yi},
  booktitle={Thirty-eighth Conference on Neural Information Processing Systems},
  year={2024},
  url={https://openreview.net/forum?id=gvlOQC6oP1}
}

Contact

If you have any questions, feel free to contact Wenhao Wang (wangwenhao0716@gmail.com).

Downloads last month
132
Edit dataset card