Overcoming Open-Set Approaches to Adversarial Defense

TMLR Paper5622 Authors

Abstract: Machine learning (ML) models are increasingly proposed to replace or augment safety-critical information processing systems, yet their fragility to evasion attacks remains a well-documented, open problem. This work analyzes a class of deep neural network defenses that add a none-of-the-above (NOTA) class as an open-set-inspired, closed-set adversarial defense. We analyze seven prominent adversarial evasion attacks developed for computer vision classification and one attack developed for natural language processing classification, identifying how these attacks fail in the presence of a NOTA defense. We use this knowledge to adapt these attacks and provide empirical evidence that adding a NOTA class alone does not solve the core challenge of defending DNNs against evasion attacks. We release our adapted attack suite to enable more rigorous future evaluations of open-set-inspired defenses.

Submission Length: Long submission (more than 12 pages of main content)

Changes Since Last Submission: N/A.

Assigned Action Editor: ~Meisam_Razaviyayn1

Submission Number: 5622