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>.

On Real-Time and Self-Taught Anomaly Detection in Optical Networks Using Hybrid Unsupervised/Supervised Learning.

Spatial Anomaly Detection in Hyperspectral Imaging Using Optical Neural Networks

Machine Learning for Real-Time Anomaly Detection in Optical Networks

Anomaly Prediction with Hybrid Supervised/Unsupervised Deep Learning for Elastic Optical Networks: A Multi-Index Correlative Approach

Machine Learning-based Anomaly Detection in Optical Fiber Monitoring

ML-based Anomaly Detection in Optical Fiber Monitoring

Label-Free Anomaly Detection Using Distributed Optical Fiber Acoustic Sensing

Distributed Online One-Class Support Vector Machine for Anomaly Detection over Networks

A task of anomaly detection for a smart satellite Internet of things system

A network anomaly detection algorithm based on semi-supervised learning and adaptive multiclass balancing

Automating Optical Network Fault Management with Machine Learning

Network Anomaly Detection and Localization

Self-Supervised Learning for Online Anomaly Detection in High-Dimensional Data Streams

A multilevel hybrid anomaly detection scheme for industrial wireless sensor networks

Dirty-data-based Alarm Prediction in Self-Optimizing Large-Scale Optical Networks

A Novel Self-Supervised Learning-Based Anomalous Node Detection Method Based on an Autoencoder for Wireless Sensor Networks

NOLSAGAN: A NO-Label Self-Attention Segmentation Method Based on Feature Reconstruction Using Generative Adversarial Networks for Optical Fiber End-Face

MSCA: an Unsupervised Anomaly Detection System for Network Security in Backbone Network.

Fault Monitoring in Passive Optical Networks using Machine Learning Techniques

AnomalyNCD: Towards Novel Anomaly Class Discovery in Industrial Scenarios

A hybrid unsupervised clustering-based anomaly detection method