5 papers
Shared Vulnerabilities in Robustness-Optimized Defenses: One Breach Exposes the Family
Hanrui Wang, Ruihao Zheng, Shuo Wang +3
Adversarial robustness optimization aims to preserve correct prediction under adversarial perturbations, and has produced substantial robustness gains through methods such as adver…
EditSleuth: A Dataset of Grounded Reasoning Chains for Image-Edit Forensics
Van-Loc Nguyen, AprilPyone MaungMaung, Minh-Triet Tran +1
Forensic analysis of AI-edited images requires more than binary real-versus-fake prediction: a useful system should localize the edit, identify its semantic type, and ground its de…
Imitation Game for Adversarial Disillusion with Chain-of-Thought Reasoning in Generative AI
Ching-Chun Chang, Fan-Yun Chen, Shih-Hong Gu +3
As the cornerstone of artificial intelligence, machine perception confronts a fundamental threat posed by adversarial illusions. These adversarial attacks manifest in two primary f…
Multimodal Adversarial Defense for Vision-Language Models by Leveraging One-To-Many Relationships
Futa Waseda, Antonio Tejero-de-Pablos, Isao Echizen
Pre-trained vision-language (VL) models are highly vulnerable to adversarial attacks. However, existing defense methods primarily focus on image classification, overlooking two key…
Rethinking Invariance Regularization in Adversarial Training to Improve Robustness-Accuracy Trade-off
Futa Waseda, Ching-Chun Chang, Isao Echizen
Adversarial training often suffers from a robustness-accuracy trade-off, where achieving high robustness comes at the cost of accuracy. One approach to mitigate this trade-off is l…