THAT: Two Head Adversarial Training for Improving Robustness at Scale

Abstract

Many variants of adversarial training have been proposed, with most research focusing on problems with relatively few classes. In this paper, we propose Two Head Adversarial Training (THAT), a two-stream adversarial learning network that is designed to handle the large-scale many-class ImageNet dataset. The proposed method trains a network with two heads and two loss functions; one to minimize feature-space domain shift between natural and adversarial images, and one to promote high classification accuracy. This combination delivers a hardened network that achieves state of the art robust accuracy while maintaining high natural accuracy on ImageNet. Through extensive experiments, we demonstrate that the proposed framework outperforms alternative methods under both standard and "free" adversarial training settings.

References in corpus (10)

• Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
• Improved Baselines with Momentum Contrastive Learning
• Fast is better than free: Revisiting adversarial training
• Adversarial Robustness through Local Linearization
• Adversarial Self-Supervised Contrastive Learning
• Evaluating and Understanding the Robustness of Adversarial Logit Pairing
• Robust Pre-Training by Adversarial Contrastive Learning
• Instance adaptive adversarial training: Improved accuracy tradeoffs in neural nets
• Metric Learning for Adversarial Robustness
• Contrastive Learning with Adversarial Examples