Abstract
Recently, Wong et al. showed that adversarial training with single-step FGSM leads to a characteristic failure mode named catastrophic overfitting (CO), in which a model becomes suddenly vulnerable to multi-step attacks. They showed that adding a random perturbation prior to FGSM (RS-FGSM) seemed to be sufficient to prevent CO. However, Andriushchenko and Flammarion observed that RS-FGSM still leads to CO for larger perturbations, and proposed an expensive regularizer (GradAlign) to avoid CO. In this work, we methodically revisit the role of noise and clipping in single-step adversarial training. Contrary to previous intuitions, we find that using a stronger noise around the clean sample combined with not clipping is highly effective in avoiding CO for large perturbation radii. Based on these observations, we then propose Noise-FGSM (N-FGSM) that, while providing the benefits of single-step adversarial training, does not suffer from CO. Empirical analyses on a large suite of experiments show that N-FGSM is able to match or surpass the performance of previous single-step methods while achieving a 3× speed-up.
Adel Bibi
Senior Researcher
My research interests include machine learning, computer vision, and optimization.