MVTec AD — A Comprehensive Real-World Dataset for Unsupervised Anomaly Detection

TLDR

Anomaly detection in natural images is critical, yet effective unsupervised methods require suitable datasets for training and evaluation. The authors present MVTec AD, a dataset of 5,354 high‑resolution images spanning multiple object and texture categories, to support unsupervised anomaly detection research. The dataset includes defect‑free training images and test images with over 70 defect types, each annotated with pixel‑precise ground truth, and the authors benchmark several deep and classical unsupervised anomaly detection approaches on it. Initial benchmarks show substantial performance gaps, underscoring the need for better methods, and the dataset is the first comprehensive, multi‑object, multi‑defect resource with pixel‑accurate annotations for real‑world anomaly detection.

Abstract

The detection of anomalous structures in natural image data is of utmost importance for numerous tasks in the field of computer vision. The development of methods for unsupervised anomaly detection requires data on which to train and evaluate new approaches and ideas. We introduce the MVTec Anomaly Detection (MVTec AD) dataset containing 5354 high-resolution color images of different object and texture categories. It 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. In addition, we provide pixel-precise ground truth regions for all anomalies. We also conduct a thorough evaluation of current state-of-the-art unsupervised anomaly detection methods based on deep architectures such as convolutional autoencoders, generative adversarial networks, and feature descriptors using pre-trained convolutional neural networks, as well as classical computer vision methods. This initial benchmark indicates that there is considerable room for improvement. To the best of our knowledge, this is the first comprehensive, multi-object, multi-defect dataset for anomaly detection that provides pixel-accurate ground truth regions and focuses on real-world applications.