Adversarial Examples Detection and Analysis with Layer-wise Autoencoders
Machine Learning
2020-06-18 v1 Cryptography and Security
Machine Learning
Abstract
We present a mechanism for detecting adversarial examples based on data representations taken from the hidden layers of the target network. For this purpose, we train individual autoencoders at intermediate layers of the target network. This allows us to describe the manifold of true data and, in consequence, decide whether a given example has the same characteristics as true data. It also gives us insight into the behavior of adversarial examples and their flow through the layers of a deep neural network. Experimental results show that our method outperforms the state of the art in supervised and unsupervised settings.
Cite
@article{arxiv.2006.10013,
title = {Adversarial Examples Detection and Analysis with Layer-wise Autoencoders},
author = {Bartosz Wójcik and Paweł Morawiecki and Marek Śmieja and Tomasz Krzyżek and Przemysław Spurek and Jacek Tabor},
journal= {arXiv preprint arXiv:2006.10013},
year = {2020}
}