Towards Robust Deep Learning with Ensemble Networks and Noisy Layers

In this paper we provide an approach for deep learning that protects against adversarial examples in image classification-type networks. The approach relies