Autoencoder Anomaly Detection Using PyTorch Dr. James McCaffrey of Microsoft Research provides full code and step-by-step examples of anomaly detection, used to find items in a dataset that are ...

We propose an unsupervised method for detecting adversarial attacks in inner layers of autoencoder (AE) networks by maximizing a non-parametric measure of anomalous node activations.

In a research paper, Baidu details AdvBox, an adversarial toolbox for testing the robustness of AI and machine learning models.