Metraipolis › archive › dc_adbd8d44a453
StyleAT: Defending Face Recognition Against Semantic Attacks — arXiv:2609.23596
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.23596
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-09T07:01:18.129Z
- SHA-256 of the archived text
52e7310cb0d7a60f80ba52b125e414551c062813480658b346d7a241ea6f55fc- Stable URL of this page
https://metraipolis.com/documento/dc_adbd8d44a453
How to cite
StyleAT: Defending Face Recognition Against Semantic Attacks — arXiv:2609.23596 — arXiv (en), https://arxiv.org/abs/2609.23596 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_adbd8d44a453 (Metraipolis archive dc_adbd8d44a453, sha256 52e7310cb0d7a60f80ba52b125e414551c062813480658b346d7a241ea6f55fc)
BibTeX
@misc{dcadbd8d44a453,
title = {StyleAT: Defending Face Recognition Against Semantic Attacks — arXiv:2609.23596},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.23596},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_adbd8d44a453; sha256 52e7310cb0d7a60f80ba52b125e414551c062813480658b346d7a241ea6f55fc},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.23596. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
With face-recognition models now embedded in everyday authentication and surveillance, recent works have pinpointed a critical weakness: these models remain acutely vulnerable to adversarial semantic edits. I.e., adversarially produced semantic alterations to the input, such as slight aging or pose changes, can induce misclassifications. Certain existing attacks are powerful, but they can be computationally costly, rendering them inadequate for developing defenses (e.g., through adversarial training). To fill the gap, we introduce BoundStyle, a potent semantic attack operating in StyleGAN's rich latent space to maximize misclassification rates. Notably, BoundStyle achieves high attack success rates while being ${\sim}{\times}9.5$ faster than existing state-of-the-art attacks, making it suitable for adversarial training. Building on BoundStyle, we develop StyleAT, an efficient adversarial training scheme that incorporates low-budget attack variants yet defends against stronger and unseen semantic attacks. We evaluate on two datasets unseen during training and seven models, and find that StyleAT boosts robust accuracy against state-of-the-art attacks and outperforms common defenses in various settings.
Tema degli esploratori: fiducia digitale e crittografia (retro-catalogo arXiv: annunci del 2026-09-22 in cs.CR).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.23596 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-09T06:56:01.029Z dalla pagina https://arxiv.org/catchup/cs.CR/2026-09-22?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).