Metric Learning for Adversarial Robustness

Abstract

Deep networks are well-known to be fragile to adversarial attacks. We conduct an empirical analysis of deep representations under the state-of-the-art attack method called PGD, and find that the attack causes the internal representation to shift closer to the "false" class. Motivated by this observation, we propose to regularize the representation space under attack with metric learning to produce more robust classifiers. By carefully sampling examples for metric learning, our learned representation not only increases robustness, but also detects previously unseen adversarial samples. Quantitative experiments show improvement of robustness accuracy by up to 4% and detection efficiency by up to 6% according to Area Under Curve score over prior work. The code of our work is available at https://github.com/columbia/Metric_Learning_Adversarial_Robustness.

Cited by in corpus (31)

• Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks
• Large-Scale Adversarial Training for Vision-and-Language Representation Learning
• RobustBench: a standardized adversarial robustness benchmark
• Boosting Adversarial Training with Hypersphere Embedding
• Diversity can be Transferred: Output Diversification for White- and Black-box Attacks
• Recent Advances in Adversarial Training for Adversarial Robustness
• Adversarial Distributional Training for Robust Deep Learning
• Opportunities and Challenges in Deep Learning Adversarial Robustness: A Survey
• Exploring Memorization in Adversarial Training
• Learnable Boundary Guided Adversarial Training
• Bag of Tricks for Adversarial Training
• Online Adversarial Purification based on Self-Supervision
• Adversarial Robustness under Long-Tailed Distribution
• Robustified Domain Adaptation
• Improving Model Robustness with Latent Distribution Locally and Globally
• Live Trojan Attacks on Deep Neural Networks
• Provably Robust Metric Learning
• Understanding and Achieving Efficient Robustness with Adversarial Supervised Contrastive Learning
• Defective Convolutional Networks
• Adversarially Robust Kernel Smoothing
• Feature Distillation With Guided Adversarial Contrastive Learning
• THAT: Two Head Adversarial Training for Improving Robustness at Scale
• Structure-Preserving Progressive Low-rank Image Completion for Defending Adversarial Attacks
• Model-Agnostic Meta-Attack: Towards Reliable Evaluation of Adversarial Robustness
• Deep Repulsive Prototypes for Adversarial Robustness
• Enhanced countering adversarial attacks via input denoising and feature restoring
• Testing Machine Translation via Referential Transparency
• Where Does the Robustness Come from? A Study of the Transformation-based Ensemble Defence
• Semantics-Preserving Adversarial Training
• Training Provably Robust Models by Polyhedral Envelope Regularization
• Identifying Layers Susceptible to Adversarial Attacks