We study adversarial robustness of neural networks from a margin maximization perspective, where margins are defined as the distances from inputs to a classifier's decision …
While adversarial training has become the de facto approach for training robust classifiers, it leads to a drop in accuracy. This has led to prior works postulating that accuracy is …
D Stutz, M Hein, B Schiele - Proceedings of the IEEE/CVF …, 2021 - openaccess.thecvf.com
Adversarial training (AT) has become the de-facto standard to obtain models robust against adversarial examples. However, AT exhibits severe robust overfitting: cross-entropy loss on …
Z Allen-Zhu, Y Li - 2021 IEEE 62nd Annual Symposium on …, 2022 - ieeexplore.ieee.org
Despite the empirical success of using adversarial training to defend deep learning models against adversarial perturbations, so far, it still remains rather unclear what the principles are …
D Stutz, M Hein, B Schiele - Proceedings of the IEEE/CVF …, 2019 - openaccess.thecvf.com
Obtaining deep networks that are robust against adversarial examples and generalize well is an open problem. A recent hypothesis even states that both robust and accurate models …
CK Mummadi, T Brox… - Proceedings of the IEEE …, 2019 - openaccess.thecvf.com
Classifiers such as deep neural networks have been shown to be vulnerable against adversarial perturbations on problems with high-dimensional input space. While adversarial …
The goal of this paper is to analyze an intriguing phenomenon recently discovered in deep networks, that is their instability to adversarial perturbations (Szegedy et al., 2014). We …
We show that there may exist an inherent tension between the goal of adversarial robustness and that of standard generalization. Specifically, training robust models may not …