Gaussian Image Anomaly Detection with Greedy Eigencomponent Selection

TL;DR

This paper presents a greedy eigencomponent selection method for image anomaly detection that outperforms PCA and NPCA, improving detection accuracy with fewer components using pre-trained CNN features.

Contribution

It introduces a novel greedy tree search approach for optimal eigencomponent selection in anomaly detection, enhancing performance over traditional methods.

Findings

Our method surpasses PCA and NPCA in detection accuracy.

Fewer components are needed to achieve high performance.

The approach is effective across different anomaly types and training data sizes.

Abstract

Anomaly detection (AD) in images, identifying significant deviations from normality, is a critical issue in computer vision. This paper introduces a novel approach to dimensionality reduction for AD using pre-trained convolutional neural network (CNN) that incorporate EfficientNet models. We investigate the importance of component selection and propose two types of tree search approaches, both employing a greedy strategy, for optimal eigencomponent selection. Our study conducts three main experiments to evaluate the effectiveness of our approach. The first experiment explores the influence of test set performance on component choice, the second experiment examines the performance when we train on one anomaly type and evaluate on all other types, and the third experiment investigates the impact of using a minimum number of images for training and selecting them based on anomaly types. Our approach aims to find the optimal subset of components that deliver the highest performance score, instead of focusing solely on the proportion of variance explained by each component and also understand the components behaviour in different settings. Our results indicate that the proposed method surpasses both Principal Component Analysis (PCA) and Negated Principal Component Analysis (NPCA) in terms of detection accuracy, even when using fewer components. Thus, our approach provides a promising alternative to conventional dimensionality reduction techniques in AD, and holds potential to enhance the efficiency and effectiveness of AD systems.