Adversarial Attack on Attackers: Post-Process to Mitigate Black-Box Score-Based Query Attacks
This addresses a practical security threat for machine learning models by defending against black-box attacks without degrading clean accuracy, though it is incremental as it builds on existing defense concepts.
The paper tackles the problem of score-based query attacks (SQAs) on deep neural networks by proposing a defense called Adversarial Attack on Attackers (AAA), which slightly modifies output logits to mislead attackers, resulting in improved accuracy under attacks, such as securing 80.59% accuracy on CIFAR-10 with WideResNet-28 compared to 67.44% for prior defenses.
The score-based query attacks (SQAs) pose practical threats to deep neural networks by crafting adversarial perturbations within dozens of queries, only using the model's output scores. Nonetheless, we note that if the loss trend of the outputs is slightly perturbed, SQAs could be easily misled and thereby become much less effective. Following this idea, we propose a novel defense, namely Adversarial Attack on Attackers (AAA), to confound SQAs towards incorrect attack directions by slightly modifying the output logits. In this way, (1) SQAs are prevented regardless of the model's worst-case robustness; (2) the original model predictions are hardly changed, i.e., no degradation on clean accuracy; (3) the calibration of confidence scores can be improved simultaneously. Extensive experiments are provided to verify the above advantages. For example, by setting $\ell_\infty=8/255$ on CIFAR-10, our proposed AAA helps WideResNet-28 secure 80.59% accuracy under Square attack (2500 queries), while the best prior defense (i.e., adversarial training) only attains 67.44%. Since AAA attacks SQA's general greedy strategy, such advantages of AAA over 8 defenses can be consistently observed on 8 CIFAR-10/ImageNet models under 6 SQAs, using different attack targets, bounds, norms, losses, and strategies. Moreover, AAA calibrates better without hurting the accuracy. Our code is available at https://github.com/Sizhe-Chen/AAA.