The official code for CMUA-Watermark: A Cross-Model Universal Adversarial Watermark for Combating Deepfakes (AAAI2022) arxiv. It is based on disrupting-deepfakes .

Contact us with huanghao@stu.pku.edu.cn, wyt@pku.edu.cn.

We will release our code soon (no later than December 31, 2021).

CMUA-Watermark is a cross-model universal adversarial watermark that can combat multiple deepfake models while protecting a myriad of facial images. With the proposed perturbation fusion strategies and automatic step size tuning, CMUA-Watermark achieves excellent protection capabilities for facial images against four face modification models (StarGAN, AttGAN, AGGAN, HiSD).

Figure 2. The quantitative results of CMUA-Watermark.

1. Prepare the Environment

   • install the lib by pip (recommend)

     pip install -r requirements.txt

2. Prepare the Datasets

   • download the CelebA datasets:

     cd stargan
     bash download.sh celeba
     

     make sure your floder (e.g. celeba_data) has img_align_celeba and list_attr_celeba.txt.
   • create the link

     ln -s your_path_to_celeba_data ./data
     

3. Prepare the Model Weights

   For your convenient usage, we prepare the weights download link in Baiduyun disk: https://pan.baidu.com/s/1GKMob15-xXpZFZTFp6fxqQ?pwd=j662

   Extract code: j662

   You can first download the weights. Then move the weight files to different floders of different models:

   cd CMUA-Watermark
   # make sure **weights** in this path.
   # If the paths bellow are not exist, please create the path (e.g., mkdir -p ./stargan/stargan_celeba_256/models).
   mv ./weights/stargan/* ./stargan/stargan_celeba_256/models
   mv ./weights/AttentionGAN/* ./AttentionGAN/AttentionGAN_v1_multi/checkpoints/celeba_256_pretrained
   mv ./weights/HiSD/* ./HiSD
   mv ./weights/AttGAN/* ./AttGAN/output/256_shortcut1_inject0_none/checkpoint

   ATTENTION! The copyright of these weight files belongs to their owners. You needs authorization for commerce, please contact to their owners!

4. Prepare the CMUA-Watermark (only for inference)

   We prepare a CMUA-Watermark for you to test its performance in Baiduyun disk: https://pan.baidu.com/s/1GKMob15-xXpZFZTFp6fxqQ?pwd=j662

   Extract code: j662

# inference in CelebA datasets with 20 images (you can change the test number in evaluate.py)
python3 universal_attack_inference.py

# inference with your own image (one image)
python3 universal_attack_inference_one_image.py ./demo_input.png # you can change the path with your own image

If your want to try your onw idea, you may need to modify the nni_config.yaml and search_space.json. These two files are the configs of NNI-based search. Thanks to the NNI, you can obtain the visualized results in your browser.

nnictl create --config ./nni_config.yaml 

Once you get the best step sizes, you need to modify the default step sizes in setting.json. It must be easy for a smart person like you~

After that,

python universal_attack.py

If you use our code / perturbation, please consider to cite our paper: CMUA-Watermark: A Cross-Model Universal Adversarial Watermark for Combating Deepfakes.

@inproceedings{huang2022cmua,
  title={Cmua-watermark: A cross-model universal adversarial watermark for combating deepfakes},
  author={Huang, Hao and Wang, Yongtao and Chen, Zhaoyu and Zhang, Yuze and Li, Yuheng and Tang, Zhi and Chu, Wei and Chen, Jingdong and Lin, Weisi and Ma, Kai-Kuang},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={36},
  number={1},
  pages={989--997},
  year={2022}
}

The project is only free for academic research purposes, but needs authorization for commerce. For commerce permission, please contact wyt@pku.edu.cn.

We use code from StarGAN, GANimation, pix2pixHD, CycleGAN, advertorch, disrupting-deepfakes and nni. These are all great repositories and we encourage you to check them out and cite them in your work.