AnomalyHop: An SSL-based Image Anomaly Localization Method

TL;DR

AnomalyHop is a transparent, fast, and effective SSL-based method for image anomaly localization that achieves high ROC-AUC performance on benchmark datasets.

Contribution

It introduces a novel SSL-based framework for anomaly localization that is mathematically transparent, easy to train, and competitive with deep learning methods.

Findings

Achieves 95.9% ROC-AUC on MVTec AD dataset

Faster inference compared to DNN-based methods

Mathematically transparent and easy to train

Abstract

An image anomaly localization method based on the successive subspace learning (SSL) framework, called AnomalyHop, is proposed in this work. AnomalyHop consists of three modules: 1) feature extraction via successive subspace learning (SSL), 2) normality feature distributions modeling via Gaussian models, and 3) anomaly map generation and fusion. Comparing with state-of-the-art image anomaly localization methods based on deep neural networks (DNNs), AnomalyHop is mathematically transparent, easy to train, and fast in its inference speed. Besides, its area under the ROC curve (ROC-AUC) performance on the MVTec AD dataset is 95.9%, which is among the best of several benchmarking methods. Our codes are publicly available at Github.