Robustness Evaluation Using Local Substitute Networks

Aleksei Kuvshinov, Crisitan Pavel, Stephan GünnemannICLR 2023 · OpenReview

Abstract

Robustness of a neural network against adversarial examples is an important topic when a deep classifier is applied in safety critical use cases like health care or autonomous driving. In order to assess the robustness, practitioners use a range of different tools ranging from the adversarial attacks to exact computation of the distance to the decision boundary. We use the fact that robustness of a neural network is a local property and empirically show that computing the same metrics for the smaller local substitute networks yields good estimates of the robustness for lower cost. To construct the substitute network we develop two pruning techniques that preserve the local properties of the initial network around a given anchor point. Our experiments on CIFAR10 and MNIST datasets prove that this approach saves a significant amount of computing time and is especially beneficial for the larger models.
BibTeX
@inproceedings{kuvshinov2023robustness,
  author = {Aleksei Kuvshinov and Crisitan Pavel and Stephan Günnemann},
  title = {{Robustness Evaluation Using Local Substitute Networks}},
  booktitle = {International Conference on Learning Representations},
  year = {2023},
  url = {https://openreview.net/forum?id=yPMsKyrn5A-},
}

Loading related content...

Figure