Improving Transformation-based Defenses against Adversarial Examples with First-order Perturbations (2103.04565v3)

Abstract: Deep neural networks have been successfully applied in various machine learning tasks. However, studies show that neural networks are susceptible to adversarial attacks. This exposes a potential threat to neural network-based intelligent systems. We observe that the probability of the correct result outputted by the neural network increases by applying small first-order perturbations generated for non-predicted class labels to adversarial examples. Based on this observation, we propose a method for counteracting adversarial perturbations to improve adversarial robustness. In the proposed method, we randomly select a number of class labels and generate small first-order perturbations for these selected labels. The generated perturbations are added together and then clamped onto a specified space. The obtained perturbation is finally added to the adversarial example to counteract the adversarial perturbation contained in the example. The proposed method is applied at inference time and does not require retraining or finetuning the model. We experimentally validate the proposed method on CIFAR-10 and CIFAR-100. The results demonstrate that our method effectively improves the defense performance of several transformation-based defense methods, especially against strong adversarial examples generated using more iterations.