De-AntiFake: Rethinking the Protective Perturbations Against Voice Cloning Attacks

Paper: https://huggingface.co/papers/2507.02606

Code and usage instructions: https://github.com/cyberrrange/De-AntiFake

Project Page: https://de-antifake.github.io

Download Model Weights

To use our model, you can follow the instruction in our github reposity. The model weights are hosted in this repository. You can download them using the huggingface_hub library:

from huggingface_hub import hf_hub_download


purification_weights_path = hf_hub_download(
    repo_id="cyberrrange/De-AntiFake",
    filename="purification.pkl"
)

print(f"Purification Model weights downloaded to: {purification_weights_path}")

refinement_weights_path = hf_hub_download(
    repo_id="cyberrrange/De-AntiFake",
    filename="refinement.ckpt"
)

print(f"Refinement Model weights downloaded to: {refinement_weights_path}")

Citation

If you use this model, please consider citing our paper:

@inproceedings{de-antifake-icml2025,
  title = {De-AntiFake: Rethinking the Protective Perturbations Against Voice Cloning Attacks},
  author = {Fan, Wei and Chen, Kejiang and Liu, Chang and Zhang, Weiming and Yu, Nenghai},
  booktitle = {International Conference on Machine Learning},
  year = {2025},
}

Contact

Technical Questions: [email protected]

General Inquiries: [email protected]