Adversarial Machine Learning at Scale

Abstract

Adversarial examples are malicious inputs designed to fool machine learning models. They often transfer from one model to another, allowing attackers to mount black box attacks without knowledge of the target model's parameters. Adversarial training is the process of explicitly training a model on adversarial examples, in order to make it more robust to attack or to reduce its test error on clean inputs. So far, adversarial training has primarily been applied to small problems. In this research, we apply adversarial training to ImageNet. Our contributions include: (1) recommendations for how to succesfully scale adversarial training to large models and datasets, (2) the observation that adversarial training confers robustness to single-step attack methods, (3) the finding that multi-step attack methods are somewhat less transferable than single-step attack methods, so single-step attacks are the best for mounting black-box attacks, and (4) resolution of a "label leaking" effect that causes adversarially trained models to perform better on adversarial examples than on clean examples, because the adversarial example construction process uses the true label and the model can learn to exploit regularities in the construction process.

17 pages, 5 figures

Cited by in corpus (15)

• Countering Adversarial Images using Input Transformations
• Measuring the tendency of CNNs to Learn Surface Statistical Regularities
• Fast Feature Fool: A data independent approach to universal adversarial perturbations
• Towards Interpretable Deep Neural Networks by Leveraging Adversarial Examples
• Optimization and Abstraction: A Synergistic Approach for Analyzing Neural Network Robustness
• Evading Defenses to Transferable Adversarial Examples by Translation-Invariant Attacks
• Detecting Adversarial Attacks on Neural Network Policies with Visual Foresight
• MaskDGA: A Black-box Evasion Technique Against DGA Classifiers and Adversarial Defenses
• Attacking Binarized Neural Networks
• Better the Devil you Know: An Analysis of Evasion Attacks using Out-of-Distribution Adversarial Examples
• Robustness Certificates Against Adversarial Examples for ReLU Networks
• Federated AI lets a team imagine together: Federated Learning of GANs
• Adversarial Defense Through Network Profiling Based Path Extraction
• Non-Determinism in Neural Networks for Adversarial Robustness
• Strategies to architect AI Safety: Defense to guard AI from Adversaries