A2: Efficient Automated Attacker for Boosting Adversarial Training
This work addresses the need for more efficient and effective adversarial training methods in machine learning security, though it appears incremental as it builds on existing AT variants.
The paper tackles the problem of generating stronger perturbations efficiently for adversarial training (AT) to improve model robustness, proposing an automated attacker called A2 that achieves reliable robustness improvements across datasets with low extra cost.
Based on the significant improvement of model robustness by AT (Adversarial Training), various variants have been proposed to further boost the performance. Well-recognized methods have focused on different components of AT (e.g., designing loss functions and leveraging additional unlabeled data). It is generally accepted that stronger perturbations yield more robust models. However, how to generate stronger perturbations efficiently is still missed. In this paper, we propose an efficient automated attacker called A2 to boost AT by generating the optimal perturbations on-the-fly during training. A2 is a parameterized automated attacker to search in the attacker space for the best attacker against the defense model and examples. Extensive experiments across different datasets demonstrate that A2 generates stronger perturbations with low extra cost and reliably improves the robustness of various AT methods against different attacks.