Regularized Training and Tight Certification for Randomized Smoothed Classifier with Provable Robustness

Abstract

Recently smoothing deep neural network based classifiers via isotropic Gaussian perturbation is shown to be an effective and scalable way to provide state-of-the-art probabilistic robustness guarantee against ${\ell }_2$ norm bounded adversarial perturbations. However, how to train a good base classifier that is accurate and robust when smoothed has not been fully investigated. In this work, we derive a new regularized risk, in which the regularizer can adaptively encourage the accuracy and robustness of the smoothed counterpart when training the base classifier. It is computationally efficient and can be implemented in parallel with other empirical defense methods. We discuss how to implement it under both standard (non-adversarial) and adversarial training scheme. At the same time, we also design a new certification algorithm, which can leverage the regularization effect to provide tighter robustness lower bound that holds with high probability. Our extensive experimentation demonstrates the effectiveness of the proposed training and certification approaches on CIFAR-10 and ImageNet datasets.

AAAI2020

References in corpus (7)

• Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examples
• Theoretically Principled Trade-off between Robustness and Accuracy
• An approach to reachability analysis for feed-forward ReLU neural networks
• Semidefinite relaxations for certifying robustness to adversarial examples
• Parseval Networks: Improving Robustness to Adversarial Examples
• Sorting out Lipschitz function approximation
• Deep Neural Networks as 0-1 Mixed Integer Linear Programs: A Feasibility Study