3

New Paradigm of Adversarial Training: Breaking Inherent Trade-Off between Accuracy and Robustness via Dummy Classes

Adversarial Training (AT) is recognized as one of the most effective methods to enhance the robustness of Deep Neural Networks (DNNs). …

Yanyun Wang, Li Liu, Zi Liang, Yi R. (May) Fung, Qingqing Ye, Haibo Hu