Paper "adversarial attack methods" Papers

PROSAC: Provably Safe Certification for Machine Learning Models under Adversarial Attacks

Chen Feng, Ziquan Liu, Zhuo Zhi et al.