Certifying Adapters: Enabling and Enhancing the Certification of Classifier Adversarial Robustness

arXiv:2405.16036v1 Announce Type: cross
Abstract: Randomized smoothing has become a leading method for achieving certified robustness in deep classifiers against l_{p}-norm adversarial perturbations. Current approaches for achieving certified robustness, such as data augmentation with Gaussian noise and adversarial training, require expensive training procedures that tune large models for different Gaussian noise levels and thus cannot leverage high-performance pre-trained neural networks. In this work, we introduce a novel certifying adapters framework (CAF) that enables and enhances the certification of classifier adversarial …

adapters adversarial arxiv augmentation certification certified cs.cr cs.cv cs.lg current data large noise procedures robustness training