Augmenting Model Robustness with Transformation-Invariant Attacks

Abstract

The vulnerability of neural networks under adversarial attacks has raised serious concerns and motivated extensive research. It has been shown that both neural networks and adversarial attacks against them can be sensitive to input transformations such as linear translation and rotation, and that human vision, which is robust against adversarial attacks, is invariant to natural input transformations. Based on these, this paper tests the hypothesis that model robustness can be further improved when it is adversarially trained against transformed attacks and transformation-invariant attacks. Experiments on MNIST, CIFAR-10, and restricted ImageNet show that while transformations of attacks alone do not affect robustness, transformation-invariant attacks can improve model robustness by 2.5\% on MNIST, 3.7\% on CIFAR-10, and 1.1\% on restricted ImageNet. We discuss the intuition behind this phenomenon.

References in corpus (18)

• Explaining and Harnessing Adversarial Examples
• Feature Squeezing: Detecting Adversarial Examples in Deep Neural Networks
• Transferability in Machine Learning: from Phenomena to Black-Box Attacks using Adversarial Samples
• Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examples
• Theoretically Principled Trade-off between Robustness and Accuracy
• Synthesizing Robust Adversarial Examples
• Towards Deep Neural Network Architectures Robust to Adversarial Examples
• Adversarially Robust Generalization Requires More Data
• Robust Physical-World Attacks on Deep Learning Models
• Countering Adversarial Images using Input Transformations
• Adversarial Risk and the Dangers of Evaluating Against Weak Attacks
• Robustness May Be at Odds with Accuracy
• Robustness and Regularization of Support Vector Machines
• Threat of Adversarial Attacks on Deep Learning in Computer Vision: A Survey
• Delving into adversarial attacks on deep policies
• Scaling provable adversarial defenses
• Shield: Fast, Practical Defense and Vaccination for Deep Learning using JPEG Compression
• Evading Defenses to Transferable Adversarial Examples by Translation-Invariant Attacks