PECAN: A Deterministic Certified Defense Against Backdoor Attacks

arXiv:2301.11824v4 Announce Type: replace
Abstract: Neural networks are vulnerable to backdoor poisoning attacks, where the attackers maliciously poison the training set and insert triggers into the test input to change the prediction of the victim model. Existing defenses for backdoor attacks either provide no formal guarantees or come with expensive-to-compute and ineffective probabilistic guarantees. We present PECAN, an efficient and certified approach for defending against backdoor attacks. The key insight powering PECAN is to apply off-the-shelf test-time evasion certification techniques …

arxiv attackers attacks backdoor backdoor attacks certified change compute cs.cr cs.lg defense defenses input networks neural networks poisoning poisoning attacks prediction test training victim vulnerable