2 papers
cs.LG2024
Soften to Defend: Towards Adversarial Robustness via Self-Guided Label Refinement
Daiwei Yu, Zhuorong Li, Lina Wei +3
Adversarial training (AT) is currently one of the most effective ways to obtain the robustness of deep neural networks against adversarial attacks. However, most AT methods suffer…
cs.LG2022
Stable and Efficient Adversarial Training through Local Linearization
Zhuorong Li, Daiwei Yu
There has been a recent surge in single-step adversarial training as it shows robustness and efficiency. However, a phenomenon referred to as ``catastrophic overfitting" has been o…