Recent Advances in Understanding Adversarial Robustness of Deep Neural Networks

TL;DR

This survey reviews recent progress in understanding adversarial robustness of deep neural networks, covering definitions, benchmarks, theoretical bounds, correlations with other indicators, and the costs of adversarial training.

Contribution

It provides a comprehensive overview of recent advances in understanding adversarial robustness from multiple perspectives, including theoretical and practical insights.

Findings

Analysis of benchmarks and theoretical bounds for robustness

Correlation between adversarial robustness and other model indicators

Discussion on the costs associated with adversarial training

Abstract

Adversarial examples are inevitable on the road of pervasive applications of deep neural networks (DNN). Imperceptible perturbations applied on natural samples can lead DNN-based classifiers to output wrong prediction with fair confidence score. It is increasingly important to obtain models with high robustness that are resistant to adversarial examples. In this paper, we survey recent advances in how to understand such intriguing property, i.e. adversarial robustness, from different perspectives. We give preliminary definitions on what adversarial attacks and robustness are. After that, we study frequently-used benchmarks and mention theoretically-proved bounds for adversarial robustness. We then provide an overview on analyzing correlations among adversarial robustness and other critical indicators of DNN models. Lastly, we introduce recent arguments on potential costs of adversarial training which have attracted wide attention from the research community.