English

Adversarial Defense via Data Dependent Activation Function and Total Variation Minimization

Machine Learning 2020-04-30 v3 Numerical Analysis Numerical Analysis Machine Learning

Abstract

We improve the robustness of Deep Neural Net (DNN) to adversarial attacks by using an interpolating function as the output activation. This data-dependent activation remarkably improves both the generalization and robustness of DNN. In the CIFAR10 benchmark, we raise the robust accuracy of the adversarially trained ResNet20 from 46%\sim 46\% to 69%\sim 69\% under the state-of-the-art Iterative Fast Gradient Sign Method (IFGSM) based adversarial attack. When we combine this data-dependent activation with total variation minimization on adversarial images and training data augmentation, we achieve an improvement in robust accuracy by 38.9%\% for ResNet56 under the strongest IFGSM attack. Furthermore, We provide an intuitive explanation of our defense by analyzing the geometry of the feature space.

Keywords

Cite

@article{arxiv.1809.08516,
  title  = {Adversarial Defense via Data Dependent Activation Function and Total Variation Minimization},
  author = {Bao Wang and Alex T. Lin and Wei Zhu and Penghang Yin and Andrea L. Bertozzi and Stanley J. Osher},
  journal= {arXiv preprint arXiv:1809.08516},
  year   = {2020}
}

Comments

17 pages, 6 figures

R2 v1 2026-06-23T04:15:05.786Z