EUVIMEDEUVIMED entdecken
Crossref

ROVER: Robust Generative Continual Identity Unlearning Against Relearning Attacks

Tairan Huang, Qiang Chen, Beibei Hu, Yunlong Zhao, Hongyan Xu, Zhiyuan Chen, Yi Chen, Xiu Su

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (7) · 5122-5130

DOI: 10.1609/aaai.v40i7.37426 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Tairan Huang und 7 weiteren Autor:innen. Die Arbeit wurde bisher 4-mal zitiert (Crossref).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Recent generative unlearning models synthesize high quality samples while protecting private information by unlearning the identity. However, existing generative identity unlearning methods face two challenges in multi-identity unlearning: 1) identity conflicts, which cause conflicts of model parameters in the continuous erasure of multiple identities; 2) fragile unlearning, where the model's unlearning ability deteriorates or fails under malicious attacks. In this paper, we introduce a critical yet under-explored task called robust multi-identity unlearning, with the goals of resolving identity conflicts to achieve interference-free unlearning and protecting against malicious attacks to achieve robust unlearning. To satisfy these goals, we propose a novel framework, RObust generatiVE continual identity unlearning against Relearning attacks (ROVER). By filtering unlearning requests with latent similarity, our method effectively isolates benign unlearning from malicious attacks to preserve identity removal integrity. Meanwhile, residual orthogonal resonator resolves identity conflicts in the continuous erasure of multiple identities, preserving stability in benign continual unlearning. Moreover, we introduce the phantom guard network to block malicious attacks by absorbing adversarial gradients, ensuring irreversible identity unlearning. The extensive experiments demonstrate that our proposed method achieves state-of-the-art performance on the task of robust multi-identity unlearning against relearning attacks.

Zitieren

APA

Huang, T., Chen, Q., Hu, B., Zhao, Y., Xu, H., Chen, Z., Chen, Y., & Su, X. (2026). ROVER: Robust Generative Continual Identity Unlearning Against Relearning Attacks. Proceedings of the AAAI Conference on Artificial Intelligence, 40(7), 5122-5130. https://doi.org/10.1609/aaai.v40i7.37426

Vancouver

Huang T, Chen Q, Hu B, Zhao Y, Xu H, Chen Z, et al. ROVER: Robust Generative Continual Identity Unlearning Against Relearning Attacks. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(7):5122-5130. doi:10.1609/aaai.v40i7.37426

BibTeX
@article{Huang2026,
  author = {Huang, Tairan and Chen, Qiang and Hu, Beibei and Zhao, Yunlong and Xu, Hongyan and Chen, Zhiyuan and Chen, Yi and Su, Xiu},
  title = {ROVER: Robust Generative Continual Identity Unlearning Against Relearning Attacks},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {7},
  pages = {5122--5130},
  doi = {10.1609/aaai.v40i7.37426}
}
Kostenlos starten →