Anomaly Multi-classification in Industrial Scenarios: Transferring Few-shot Learning to a New Task

TL;DR

This paper introduces a new anomaly multi-classification task for industrial scenarios, proposing a transfer learning approach that combines RelationNet, PatchCore, and contrastive learning, achieving superior results on benchmark datasets.

Contribution

It defines the anomaly multi-classification task and develops a baseline model with a novel data generation method and contrastive learning for effective transfer learning.

Findings

Superior performance on MvTec AD and MvTec3D AD datasets.

Effective transfer of few-shot learning to industrial anomaly classification.

Enhanced accuracy over traditional fine-tuning methods.

Abstract

In industrial scenarios, it is crucial not only to identify anomalous items but also to classify the type of anomaly. However, research on anomaly multi-classification remains largely unexplored. This paper proposes a novel and valuable research task called anomaly multi-classification. Given the challenges in applying few-shot learning to this task, due to limited training data and unique characteristics of anomaly images, we introduce a baseline model that combines RelationNet and PatchCore. We propose a data generation method that creates pseudo classes and a corresponding proxy task, aiming to bridge the gap in transferring few-shot learning to industrial scenarios. Furthermore, we utilize contrastive learning to improve the vanilla baseline, achieving much better performance than directly fine-tune a ResNet. Experiments conducted on MvTec AD and MvTec3D AD demonstrate that our approach shows superior performance in this novel task.