The robustness of neural networks in image classification is important to resist adversarial
attacks. Although many researchers proposed to enhance the network robustness by …

Abstract Stochastic Neural Networks (SNNs) have emerged as a promising tool for
improving model adversarial robustness by injecting uncertainty into model activations or …

Adversarial training (AT) is one of the most effective ways for deep neural network models to
resist adversarial examples. However, there is still a significant gap between robust training …

Prior work on Private Inference (PI)--inferences performed directly on encrypted input--has
focused on minimizing a network's ReLUs, which have been assumed to dominate PI …

Existing works have made great progress in improving adversarial robustness, but typically
test their method only on data from the same distribution as the training data, ie in …