DP-TRAE: A Dual-Phase Merging Transferable Reversible Adversarial Example for Image Privacy Protection
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909607313539072 |
|---|---|
| author | Du, Xia Zhu, Jiajie Zhou, Jizhe Pun, Chi-man Lin, Zheng Wu, Cong Chen, Zhe Luo, Jun |
| author_facet | Du, Xia Zhu, Jiajie Zhou, Jizhe Pun, Chi-man Lin, Zheng Wu, Cong Chen, Zhe Luo, Jun |
| contents | In the field of digital security, Reversible Adversarial Examples (RAE) combine adversarial attacks with reversible data hiding techniques to effectively protect sensitive data and prevent unauthorized analysis by malicious Deep Neural Networks (DNNs). However, existing RAE techniques primarily focus on white-box attacks, lacking a comprehensive evaluation of their effectiveness in black-box scenarios. This limitation impedes their broader deployment in complex, dynamic environments. Further more, traditional black-box attacks are often characterized by poor transferability and high query costs, significantly limiting their practical applicability. To address these challenges, we propose the Dual-Phase Merging Transferable Reversible Attack method, which generates highly transferable initial adversarial perturbations in a white-box model and employs a memory augmented black-box strategy to effectively mislead target mod els. Experimental results demonstrate the superiority of our approach, achieving a 99.0% attack success rate and 100% recovery rate in black-box scenarios, highlighting its robustness in privacy protection. Moreover, we successfully implemented a black-box attack on a commercial model, further substantiating the potential of this approach for practical use. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_06860 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DP-TRAE: A Dual-Phase Merging Transferable Reversible Adversarial Example for Image Privacy Protection Du, Xia Zhu, Jiajie Zhou, Jizhe Pun, Chi-man Lin, Zheng Wu, Cong Chen, Zhe Luo, Jun Cryptography and Security Artificial Intelligence Machine Learning In the field of digital security, Reversible Adversarial Examples (RAE) combine adversarial attacks with reversible data hiding techniques to effectively protect sensitive data and prevent unauthorized analysis by malicious Deep Neural Networks (DNNs). However, existing RAE techniques primarily focus on white-box attacks, lacking a comprehensive evaluation of their effectiveness in black-box scenarios. This limitation impedes their broader deployment in complex, dynamic environments. Further more, traditional black-box attacks are often characterized by poor transferability and high query costs, significantly limiting their practical applicability. To address these challenges, we propose the Dual-Phase Merging Transferable Reversible Attack method, which generates highly transferable initial adversarial perturbations in a white-box model and employs a memory augmented black-box strategy to effectively mislead target mod els. Experimental results demonstrate the superiority of our approach, achieving a 99.0% attack success rate and 100% recovery rate in black-box scenarios, highlighting its robustness in privacy protection. Moreover, we successfully implemented a black-box attack on a commercial model, further substantiating the potential of this approach for practical use. |
| title | DP-TRAE: A Dual-Phase Merging Transferable Reversible Adversarial Example for Image Privacy Protection |
| topic | Cryptography and Security Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2505.06860 |