Recent Advances in Adversarial Training for Adversarial Robustness

TL;DR

This paper systematically reviews recent progress in adversarial training for deep learning robustness, introducing a new taxonomy, discussing generalization issues, and outlining future research challenges and directions.

Contribution

It provides the first comprehensive survey with a novel taxonomy of adversarial training methods and discusses generalization problems and future challenges in the field.

Findings

Various improvements in adversarial training methods

Identification of key generalization issues

Outline of future research directions

Abstract

Adversarial training is one of the most effective approaches defending against adversarial examples for deep learning models. Unlike other defense strategies, adversarial training aims to promote the robustness of models intrinsically. During the last few years, adversarial training has been studied and discussed from various aspects. A variety of improvements and developments of adversarial training are proposed, which were, however, neglected in existing surveys. For the first time in this survey, we systematically review the recent progress on adversarial training for adversarial robustness with a novel taxonomy. Then we discuss the generalization problems in adversarial training from three perspectives. Finally, we highlight the challenges which are not fully tackled and present potential future directions.