Anomaly Detection

This repository demonstrates the implementation of anomaly detection using Anomalib's PatchCore model. It includes a configurable pipeline for training and classification, leveraging YAML-based settings and the Anomalib library for efficient anomaly detection workflows.

This repository is for testing and experimenting with [Anomalib](https://github.com/openvinotoolkit/anomalib), an open-source library for anomaly detection.

Anomaly Detection on the MVTec-AD dataset using the Anomalib library. Includes PatchCore, FastFlow, and STFPM models with AUROC evaluations by category (tile, leather, grid) and overall benchmarks.

This project focuses on anomaly detection for quality control and automated inspection, utilizing the MVTec-AD dataset. It leverages the Anomalib library to train and evaluate models like PatchCore and EfficientAD for identifying and analyzing anomalies in product images.