Provable Unrestricted Adversarial Training without Compromise with Generalizability

Adversarial training (AT) is widely considered as the most promising strategy to defend against adversarial attacks and has drawn increasing interest from researchers. However, the existing AT methods still suffer from two challenges. First, they are unable to handle unrestricted adversarial example...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	IEEE transactions on pattern analysis and machine intelligence. - 1979. - PP(2024) vom: 14. Mai
1. Verfasser:	Zhang, Lilin (VerfasserIn)
Weitere Verfasser:	Yang, Ning, Sun, Yanchao, Yu, Philip S
Format:	Online-Aufsatz
Sprache:	English
Veröffentlicht:	2024
Zugriff auf das übergeordnete Werk:	IEEE transactions on pattern analysis and machine intelligence
Schlagworte:	Journal Article