Pro-AD: Learning Comprehensive Prototypes with Prototype-based Constraint for Multi-class Unsupervised Anomaly Detection

2025-06-16

AI Summary

To address two key bottlenecks in multi-class unsupervised anomaly detection—limited representational capacity of prototypes and erroneous reconstruction of anomalies via attention mechanisms (“soft identity mapping”)—this paper proposes a novel prototype learning framework. The method introduces an extensible multi-level prototype set, a dynamic bidirectional decoder enabling both forward reconstruction and backward verification, and prototype-semantic constraint regularization to suppress anomalous responses. For the first time, it achieves comprehensive modeling of normal patterns and effective suppression of anomaly reconstruction across multi-scale feature spaces. Evaluated on mainstream benchmarks including MVTec-AD and VisA, the approach achieves state-of-the-art performance, significantly improving detection accuracy, robustness, and cross-class generalization capability.

Technology Category

• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Computer Vision: Representation Learning for Vision
• Data Mining & Knowledge Management: Anomaly/Outlier Detection

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Prototype-based reconstruction methods for unsupervised anomaly detection utilize a limited set of learnable prototypes which only aggregates insufficient normal information, resulting in undesirable reconstruction. However, increasing the number of prototypes may lead to anomalies being well reconstructed through the attention mechanism, which we refer to as the"Soft Identity Mapping"problem. In this paper, we propose Pro-AD to address these issues and fully utilize the prototypes to boost the performance of anomaly detection. Specifically, we first introduce an expanded set of learnable prototypes to provide sufficient capacity for semantic information. Then we employ a Dynamic Bidirectional Decoder which integrates the process of the normal information aggregation and the target feature reconstruction via prototypes, with the aim of allowing the prototypes to aggregate more comprehensive normal semantic information from different levels of the image features and the target feature reconstruction to not only utilize its contextual information but also dynamically leverage the learned comprehensive prototypes. Additionally, to prevent the anomalies from being well reconstructed using sufficient semantic information through the attention mechanism, Pro-AD introduces a Prototype-based Constraint that applied within the target feature reconstruction process of the decoder, which further improves the performance of our approach. Extensive experiments on multiple challenging benchmarks demonstrate that our Pro-AD achieve state-of-the-art performance, highlighting its superior robustness and practical effectiveness for Multi-class Unsupervised Anomaly Detection task.

Problem

• Addresses insufficient normal information aggregation in prototype-based anomaly detection
• Solves the 'Soft Identity Mapping' problem in multi-class unsupervised anomaly detection
• Enhances prototype utilization to improve anomaly detection performance

Innovation

• Expanded learnable prototypes for sufficient semantic information
• Dynamic Bidirectional Decoder integrates normal aggregation and reconstruction
• Prototype-based Constraint prevents anomaly reconstruction via attention