We study the recently introduced stability training as a general-purpose method to increase the robustness of deep neural networks against input perturbations. In particular, we explore its use as an alternative to data augmentation and validate its performance against a number of distortion types and transformations including adversarial examples. In our image classification experiments using ImageNet data stability training performs on a par or even outperforms data augmentation for specific transformations, while consistently offering improved robustness against a broader range of distortion strengths and types unseen during training, a considerably smaller hyperparameter dependence and less potentially negative side effects compared to data augmentation.
@article{arxiv.1906.00735,
title = {Achieving Generalizable Robustness of Deep Neural Networks by Stability Training},
author = {Jan Laermann and Wojciech Samek and Nils Strodthoff},
journal= {arXiv preprint arXiv:1906.00735},
year = {2019}
}