Abstract:In the industrial scenario, anomaly detection could locate but cannot classify anomalies. To complete their capability, we study to automatically discover and recognize visual classes of industrial anomalies. In terms of multi-class anomaly classification, previous methods cluster anomalies represented by frozen pre-trained models but often fail due to poor discrimination. Novel class discovery (NCD) has the potential to tackle this. However, it struggles with non-prominent and semantically weak anomalies that challenge network learning focus. To address these, we introduce AnomalyNCD, a multi-class anomaly classification framework compatible with existing anomaly detection methods. This framework learns anomaly-specific features and classifies anomalies in a self-supervised manner. Initially, a technique called Main Element Binarization (MEBin) is first designed, which segments primary anomaly regions into masks to alleviate the impact of incorrect detections on learning. Subsequently, we employ mask-guided contrastive representation learning to improve feature discrimination, which focuses network attention on isolated anomalous regions and reduces the confusion of erroneous inputs through re-corrected pseudo labels. Finally, to enable flexible classification at both region and image levels during inference, we develop a region merging strategy that determines the overall image category based on the classified anomaly regions. Our method outperforms the state-of-the-art works on the MVTec AD and MTD datasets. Compared with the current methods, AnomalyNCD combined with zero-shot anomaly detection method achieves a 10.8% $F_1$ gain, 8.8% NMI gain, and 9.5% ARI gain on MVTec AD, 12.8% $F_1$ gain, 5.7% NMI gain, and 10.8% ARI gain on MTD. The source code is available at <a class="link-external link-https" href="https://github.com/HUST-SLOW/AnomalyNCD" rel="external noopener nofollow">this https URL</a>.

Component-aware anomaly detection framework for adjustable and logical industrial visual inspection

SAM-LAD: Segment Anything Model Meets Zero-Shot Logic Anomaly Detection

A Diffusion-Based Framework for Multi-Class Anomaly Detection

VMAD: Visual-enhanced Multimodal Large Language Model for Zero-Shot Anomaly Detection

Exploring Large Vision-Language Models for Robust and Efficient Industrial Anomaly Detection

CSAD: Unsupervised Component Segmentation for Logical Anomaly Detection

Uni-3DAD: GAN-Inversion Aided Universal 3D Anomaly Detection on Model-free Products

Towards Continual Adaptation in Industrial Anomaly Detection

Few Shot Part Segmentation Reveals Compositional Logic for Industrial Anomaly Detection

A Comprehensive Library for Benchmarking Multi-class Visual Anomaly Detection

AnomalyNCD: Towards Novel Anomaly Class Discovery in Industrial Scenarios

MLAD: Manifest and Latent Anomaly Detection based on the Integration of Reconstruction and MLFP-KNN Methods

Cognitive visual anomaly detection with constrained latent representations for industrial inspection robot

UniVAD: A Training-free Unified Model for Few-shot Visual Anomaly Detection

Component Decomposition Analysis for Hyperspectral Anomaly Detection

Self-supervised Feature Adaptation for 3D Industrial Anomaly Detection

TAB: Text-Align Anomaly Backbone Model for Industrial Inspection Tasks

GeneralAD: Anomaly Detection Across Domains by Attending to Distorted Features

Customizing Visual-Language Foundation Models for Multi-modal Anomaly Detection and Reasoning

Anomaly Detection by Adapting a pre-trained Vision Language Model

Separating Novel Features for Logical Anomaly Detection: A Straightforward yet Effective Approach