1. Benchmarks
MVTec AD is a real-world industrial inspection tasks that contains normal, i.e., defect-free, images intended for training and images with anomalies intended for testing. The anomalies manifest themselves in the form of over 70 different types of defects such as scratches, dents, contaminations, and various structural changes.
The MVTec Anomaly Detection dataset comprises 15 categories with 3629 images for training and validation and 1725 images for testing.
To prepare the dataset for OpenOOD benchmark, please download the zip file from here and extract it into images.
