Improved robustness to adversarial examples using Lipschitz regularization of the loss

Abstract

We augment adversarial training (AT) with worst case adversarial training (WCAT) which improves adversarial robustness by 11% over the current state-of-the-art result in the ${\ell }_2$ norm on CIFAR-10. We obtain verifiable average case and worst case robustness guarantees, based on the expected and maximum values of the norm of the gradient of the loss. We interpret adversarial training as Total Variation Regularization, which is a fundamental tool in mathematical image processing, and WCAT as Lipschitz regularization.

arXiv commentMerged with arXiv:1808.09540

References in corpus (5)

• Spectral Normalization for Generative Adversarial Networks
• Improved Regularization of Convolutional Neural Networks with Cutout
• Ensemble Adversarial Training: Attacks and Defenses
• Formal Guarantees on the Robustness of a Classifier against Adversarial Manipulation
• SafetyNet: Detecting and Rejecting Adversarial Examples Robustly

Cited by in corpus (10)

• Improving Adversarial Robustness via Channel-wise Activation Suppressing
• Interpreting Adversarial Examples by Activation Promotion and Suppression
• The Geometry of Adversarial Training in Binary Classification
• The coupling effect of Lipschitz regularization in deep neural networks
• Enhancing Intrinsic Adversarial Robustness via Feature Pyramid Decoder
• Adversarial training applied to Convolutional Neural Network for photometric redshift predictions
• Coarse-Grained Smoothness for RL in Metric Spaces
• Towards Natural Robustness Against Adversarial Examples
• Nonparametric Online Learning Using Lipschitz Regularized Deep Neural Networks
• On Model Robustness Against Adversarial Examples