Achieving Generalizable Robustness of Deep Neural Networks by Stability Training
Abstract
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.
- Publication:
-
arXiv e-prints
- Pub Date:
- June 2019
- DOI:
- 10.48550/arXiv.1906.00735
- arXiv:
- arXiv:1906.00735
- Bibcode:
- 2019arXiv190600735L
- Keywords:
-
- Computer Science - Machine Learning;
- Statistics - Machine Learning
- E-Print:
- 18 pages, 25 figures