Progressive Backdoor Erasing via connecting Backdoor and Adversarial Attacks
This addresses the security issue of backdoor attacks in machine learning models, offering a novel defense method that works without requiring additional clean data, which is incremental in connecting backdoor and adversarial attacks.
The paper tackles the problem of backdoor attacks in deep neural networks by proposing a Progressive Backdoor Erasing (PBE) algorithm that leverages adversarial attacks to purify infected models, achieving effective backdoor removal without a clean extra dataset and outperforming existing defenses against five state-of-the-art attacks.
Deep neural networks (DNNs) are known to be vulnerable to both backdoor attacks as well as adversarial attacks. In the literature, these two types of attacks are commonly treated as distinct problems and solved separately, since they belong to training-time and inference-time attacks respectively. However, in this paper we find an intriguing connection between them: for a model planted with backdoors, we observe that its adversarial examples have similar behaviors as its triggered images, i.e., both activate the same subset of DNN neurons. It indicates that planting a backdoor into a model will significantly affect the model's adversarial examples. Based on these observations, a novel Progressive Backdoor Erasing (PBE) algorithm is proposed to progressively purify the infected model by leveraging untargeted adversarial attacks. Different from previous backdoor defense methods, one significant advantage of our approach is that it can erase backdoor even when the clean extra dataset is unavailable. We empirically show that, against 5 state-of-the-art backdoor attacks, our PBE can effectively erase the backdoor without obvious performance degradation on clean samples and significantly outperforms existing defense methods.