Tail of Distribution GAN (TailGAN): Generative-Adversarial-Network-Based Boundary Formation

TL;DR

TailGAN is a novel GAN-based model designed to generate and detect anomalies near the distribution's tail, improving boundary detection and out-of-distribution identification in various datasets.

Contribution

The paper introduces TailGAN, a new GAN model that effectively generates tail boundary samples for enhanced anomaly detection and out-of-distribution identification.

Findings

TailGAN effectively generates tail boundary samples.

Achieves competitive OOD detection performance.

Addresses support with disjoint components.

Abstract

Generative Adversarial Networks (GAN) are a powerful methodology and can be used for unsupervised anomaly detection, where current techniques have limitations such as the accurate detection of anomalies near the tail of a distribution. GANs generally do not guarantee the existence of a probability density and are susceptible to mode collapse, while few GANs use likelihood to reduce mode collapse. In this paper, we create a GAN-based tail formation model for anomaly detection, the Tail of distribution GAN (TailGAN), to generate samples on the tail of the data distribution and detect anomalies near the support boundary. Using TailGAN, we leverage GANs for anomaly detection and use maximum entropy regularization. Using GANs that learn the probability of the underlying distribution has advantages in improving the anomaly detection methodology by allowing us to devise a generator for boundary samples, and use this model to characterize anomalies. TailGAN addresses supports with disjoint components and achieves competitive performance on images. We evaluate TailGAN for identifying Out-of-Distribution (OoD) data and its performance evaluated on MNIST, CIFAR-10, Baggage X-Ray, and OoD data shows competitiveness compared to methods from the literature.