That's BAD: Blind Anomaly Detection by Implicit Local Feature Clustering

TL;DR

This paper introduces PatchCluster, a novel unsupervised method for blind anomaly detection that identifies anomalies in images without prior normal data, achieving competitive results with state-of-the-art one-class methods.

Contribution

PatchCluster is the first method to address blind anomaly detection by converting it into a local outlier detection problem, eliminating the need for normal data.

Findings

PatchCluster accurately detects image- and pixel-level anomalies.

It performs comparably to state-of-the-art one-class methods.

The method is effective without human-annotated normal data.

Abstract

Recent studies on visual anomaly detection (AD) of industrial objects/textures have achieved quite good performance. They consider an unsupervised setting, specifically the one-class setting, in which we assume the availability of a set of normal (\textit{i.e.}, anomaly-free) images for training. In this paper, we consider a more challenging scenario of unsupervised AD, in which we detect anomalies in a given set of images that might contain both normal and anomalous samples. The setting does not assume the availability of known normal data and thus is completely free from human annotation, which differs from the standard AD considered in recent studies. For clarity, we call the setting blind anomaly detection (BAD). We show that BAD can be converted into a local outlier detection problem and propose a novel method named PatchCluster that can accurately detect image- and pixel-level anomalies. Experimental results show that PatchCluster shows a promising performance without the knowledge of normal data, even comparable to the SOTA methods applied in the one-class setting needing it.