Despite numerous attempts to defend deep learning based image classifiers, they remain susceptible to the adversarial attacks. This paper proposes a technique to identify susceptible classes, those classes that are more easily subverted. To identify the susceptible classes we use distance-based measures and apply them on a trained model. Based on the distance among original classes, we create mapping among original classes and adversarial classes that helps to reduce the randomness of a model to a significant amount in an adversarial setting. We analyze the high dimensional geometry among the feature classes and identify the k most susceptible target classes in an adversarial attack. We conduct experiments using MNIST, Fashion MNIST, CIFAR-10 (ImageNet and ResNet-32) datasets. Finally, we evaluate our techniques in order to determine which distance-based measure works best and how the randomness of a model changes with perturbation.
@article{arxiv.1905.13284,
title = {Identifying Classes Susceptible to Adversarial Attacks},
author = {Rangeet Pan and Md Johirul Islam and Shibbir Ahmed and Hridesh Rajan},
journal= {arXiv preprint arXiv:1905.13284},
year = {2019}
}