Gradient-based Novelty Detection Boosted by Self-supervised Binary Classification

TL;DR

This paper introduces a self-supervised gradient-based method for novelty detection that does not require prior OOD data, leveraging a binary classifier to improve detection accuracy across multiple datasets.

Contribution

It proposes a novel approach combining gradient Mahalanobis distance with self-supervised binary classification for OOD detection without pre-defined OOD data.

Findings

Outperforms state-of-the-art methods on multiple datasets

Achieves higher AUROC and AUPR metrics

Effectively learns OOD classes in continual learning

Abstract

Novelty detection aims to automatically identify out-of-distribution (OOD) data, without any prior knowledge of them. It is a critical step in data monitoring, behavior analysis and other applications, helping enable continual learning in the field. Conventional methods of OOD detection perform multi-variate analysis on an ensemble of data or features, and usually resort to the supervision with OOD data to improve the accuracy. In reality, such supervision is impractical as one cannot anticipate the anomalous data. In this paper, we propose a novel, self-supervised approach that does not rely on any pre-defined OOD data: (1) The new method evaluates the Mahalanobis distance of the gradients between the in-distribution and OOD data. (2) It is assisted by a self-supervised binary classifier to guide the label selection to generate the gradients, and maximize the Mahalanobis distance. In the evaluation with multiple datasets, such as CIFAR-10, CIFAR-100, SVHN and TinyImageNet, the proposed approach consistently outperforms state-of-the-art supervised and unsupervised methods in the area under the receiver operating characteristic (AUROC) and area under the precision-recall curve (AUPR) metrics. We further demonstrate that this detector is able to accurately learn one OOD class in continual learning.