Recent Advances in Understanding Adversarial Robustness of Deep Neural Networks
arXiv:2011.01539
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.
arXiv commentInitial Draft
References in corpus (14)
- ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness
- Adversarially Robust Generalization Requires More Data
- Adversarial Examples Are Not Bugs, They Are Features
- Adversarial Attack on Graph Structured Data
- Sparse DNNs with Improved Adversarial Robustness
- Interpreting Adversarially Trained Convolutional Neural Networks
- Instance adaptive adversarial training: Improved accuracy tradeoffs in neural nets
- First-order Adversarial Vulnerability of Neural Networks and Input Dimension
- Are Accuracy and Robustness Correlated?
- Adversarial Examples in Modern Machine Learning: A Review
- Theoretical Analysis of Adversarial Learning: A Minimax Approach
- Using Videos to Evaluate Image Model Robustness
- Manifold Regularization for Locally Stable Deep Neural Networks
- On Extensions of CLEVER: A Neural Network Robustness Evaluation Algorithm