Textile Anomaly Detection: Evaluation of the State-of-the-Art for Automated Quality Inspection of Carpet

TL;DR

This paper evaluates unsupervised anomaly detection models for automated quality inspection of wool carpets, focusing on accuracy, false detections, and real-time inference performance in manufacturing settings.

Contribution

It provides a comprehensive comparison of state-of-the-art models, highlighting the effectiveness of student-teacher networks and multi-class training for carpet anomaly detection.

Findings

Student-teacher networks achieved highest accuracy and lowest false detections.

Multi-class training yielded comparable or better results than single-class.

Most models had inference times around 0.16s per image on GPU.

Abstract

In this study, state-of-the-art unsupervised detection models were evaluated for the purpose of automated anomaly inspection of wool carpets. A custom dataset of four unique types of carpet textures was created to thoroughly test the models and their robustness in detecting subtle anomalies in complex textures. Due to the requirements of an inline inspection system in a manufacturing use case, the metrics of importance in this study were accuracy in detecting anomalous areas, the number of false detections, and the inference times of each model for real-time performance. Of the evaluated models, the student-teacher network based methods were found on average to yield the highest detection accuracy and lowest false detection rates. When trained on a multi-class dataset the models were found to yield comparable if not better results than single-class training. Finally, in terms of detection speed, with exception to the generative model, all other evaluated models were found to have comparable inference times on a GPU, with an average of 0.16s per image. On a CPU, most of these models typically produced results between 1.5 to 2 times the respective GPU inference times.