Detecting Anomalies with ${{\textrm{Latent}}Out}$: Novel Scores, Architectures, and Settings

Abstract

${{\textrm{Latent}}Out}$ is a recently introduced algorithm for unsupervised anomaly detection which enhances latent space-based neural methods, namely (Variational) Autoencoders, GANomaly and ANOGan architectures. The main idea behind it is to exploit both the latent space and the baseline score of these architectures in order to provide a refined anomaly score performing density estimation in the augmented latent-space/baseline-score feature space. In this paper we extend the research on the ${{\textrm{Latent}}Out}$ methodology in three directions: first, we provide a novel score performing a different kind of density estimation at a reduced computational cost; second, we experiment the combination of ${{\textrm{Latent}}Out}$ with GAAL architectures, a novel type of Generative Adversarial Networks for unsupervised anomaly detection; third, we investigate performances of ${{\textrm{Latent}}Out}$ acting as a one-class classifier. The experiments show that all the variants of ${{\textrm{Latent}}Out}$ here introduced improve performances of the baseline methods to which they are applied, both in the unsupervised and in the semi-supervised settings.