Yang, Yang and Idris, Norisma and Liu, Chang and Wu, Hui and Yu, Dingguo (2024) A destructive active defense algorithm for deepfake face images. PeerJ Computer Science, 10. ISSN 2376-5992, DOI https://doi.org/10.7717/PEERJ-CS.2356.
Full text not available from this repository.Abstract
The harm caused by deepfake face images is increasing. To proactively defend against this threat, this paper innovatively proposes a destructive active defense algorithm for deepfake face images (DADFI). This algorithm adds slight perturbations to the original face images to generate adversarial samples. These perturbations are imperceptible to the human eye but cause significant distortions in the outputs of mainstream deepfake models. Firstly, the algorithm generates adversarial samples that maintain high visual fidelity and authenticity. Secondly, in a black-box scenario, the adversarial samples are used to attack deepfake models to enhance their offensive capabilities. Finally, destructive attack experiments were conducted on the mainstream face datasets CASIAFaceV5 and CelebA. The results demonstrate that the proposed DADFI algorithm not only improves the generation speed of adversarial samples but also increases the success rate of active defense. This achievement can effectively reduce the harm caused by deepfake face images.
| Item Type: | Article |
|---|---|
| Funders: | National Office for Philosophy and Social Sciences [Grant no. 22BSH025], National Natural Science Foundation of China [Grant no. 62206241], Key Research and Development Program of Zhejiang Province [Grant no. 2021C03138], Medium and Long-Term Science and Technology Plan for Radio [Grant no. 2022AD0400] |
| Uncontrolled Keywords: | Deep fake; Face images; Active defense; Adversarial samples |
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science |
| Divisions: | Faculty of Computer Science & Information Technology > Department of Artificial Intelligence |
| Depositing User: | Ms. Juhaida Abd Rahim |
| Date Deposited: | 27 Oct 2025 08:09 |
| Last Modified: | 27 Oct 2025 08:09 |
| URI: | http://eprints.um.edu.my/id/eprint/46389 |
Actions (login required)
 |
View Item |
