Improving Fast Adversarial Training via Self-Knowledge Guidance

TL;DR

This paper introduces SKG-FAT, a novel fast adversarial training method that adaptively optimizes training data using self-knowledge guidance to improve robustness and address class disparity and misalignment issues.

Contribution

It proposes self-knowledge guided regularization and label relaxation to enhance fast adversarial training, a novel approach addressing imbalance and misalignment in robustness training.

Findings

Outperforms state-of-the-art methods on four datasets.

Improves robustness while maintaining clean accuracy.

Effectively alleviates class disparity and misalignment issues.

Abstract

Adversarial training has achieved remarkable advancements in defending against adversarial attacks. Among them, fast adversarial training (FAT) is gaining attention for its ability to achieve competitive robustness with fewer computing resources. Existing FAT methods typically employ a uniform strategy that optimizes all training data equally without considering the influence of different examples, which leads to an imbalanced optimization. However, this imbalance remains unexplored in the field of FAT. In this paper, we conduct a comprehensive study of the imbalance issue in FAT and observe an obvious class disparity regarding their performances. This disparity could be embodied from a perspective of alignment between clean and robust accuracy. Based on the analysis, we mainly attribute the observed misalignment and disparity to the imbalanced optimization in FAT, which motivates us to optimize different training data adaptively to enhance robustness. Specifically, we take disparity and misalignment into consideration. First, we introduce self-knowledge guided regularization, which assigns differentiated regularization weights to each class based on its training state, alleviating class disparity. Additionally, we propose self-knowledge guided label relaxation, which adjusts label relaxation according to the training accuracy, alleviating the misalignment and improving robustness. By combining these methods, we formulate the Self-Knowledge Guided FAT (SKG-FAT), leveraging naturally generated knowledge during training to enhance the adversarial robustness without compromising training efficiency. Extensive experiments on four standard datasets demonstrate that the SKG-FAT improves the robustness and preserves competitive clean accuracy, outperforming the state-of-the-art methods.