A Recover-then-Discriminate Framework for Robust Anomaly Detection

TL;DR

This paper introduces a Recover-then-Discriminate framework for anomaly detection, addressing key failure cases by leveraging feature-level discrimination and self-generated feature maps, resulting in state-of-the-art accuracy.

Contribution

The novel ReDi framework combines recovery and discrimination strategies, improving anomaly detection accuracy over existing methods.

Findings

ReDi achieves state-of-the-art results on benchmark datasets.

Feature-level discrimination enhances abnormality recognition.

Recovery of essential features improves detection robustness.

Abstract

Anomaly detection (AD) has been extensively studied and applied in a wide range of scenarios in the recent past. However, there are still gaps between achieved and desirable levels of recognition accuracy for making AD for practical applications. In this paper, we start from an insightful analysis of two types of fundamental yet representative failure cases in the baseline model, and reveal reasons that hinder current AD methods from achieving a higher recognition accuracy. Specifically, by Case-1, we found that the main reasons detrimental to current AD methods is that the inputs to the recovery model contain a large number of detailed features to be recovered, which leads to the normal/abnormal area has-not/has been recovered into its original state. By Case-2, we surprisingly found that the abnormal area that cannot be recognized in image-level representations can be easily recognized in the feature-level representation. Based on the above observations, we propose a novel Recover-then-Discriminate (ReDi) framework for AD. ReDi takes a self-generated feature map and a selected prompted image as explicit input information to solve problems in case-1. Concurrently, a feature-level discriminative network is proposed to enhance abnormal differences between the recovered representation and the input representation. Extensive experimental results on two popular yet challenging AD datasets validate that ReDi achieves the new state-of-the-art accuracy.