Center-aware Residual Anomaly Synthesis for Multi-class Industrial Anomaly Detection

TL;DR

The paper introduces CRAS, a novel multi-class industrial anomaly detection method that reduces interference and overlap issues by center-aware residual learning and adaptive noise adjustment, achieving high accuracy and efficiency.

Contribution

CRAS is the first unified multi-class anomaly detection approach that effectively mitigates inter-class interference and intra-class overlap through innovative center-aware residual learning.

Findings

Superior detection accuracy on diverse datasets

Competitive inference speed

Effective handling of inter-class interference and intra-class overlap

Abstract

Anomaly detection plays a vital role in the inspection of industrial images. Most existing methods require separate models for each category, resulting in multiplied deployment costs. This highlights the challenge of developing a unified model for multi-class anomaly detection. However, the significant increase in inter-class interference leads to severe missed detections. Furthermore, the intra-class overlap between normal and abnormal samples, particularly in synthesis-based methods, cannot be ignored and may lead to over-detection. To tackle these issues, we propose a novel Center-aware Residual Anomaly Synthesis (CRAS) method for multi-class anomaly detection. CRAS leverages center-aware residual learning to couple samples from different categories into a unified center, mitigating the effects of inter-class interference. To further reduce intra-class overlap, CRAS introduces distance-guided anomaly synthesis that adaptively adjusts noise variance based on normal data distribution. Experimental results on diverse datasets and real-world industrial applications demonstrate the superior detection accuracy and competitive inference speed of CRAS. The source code and the newly constructed dataset are publicly available at https://github.com/cqylunlun/CRAS.