The MVTec 3D-AD Dataset for Unsupervised 3D Anomaly Detection and Localization

TL;DR

This paper introduces the MVTec 3D-AD dataset, a comprehensive collection of 3D scans designed for unsupervised anomaly detection and localization in manufacturing, highlighting the challenges and potential for advancing 3D defect detection methods.

Contribution

It provides the first extensive 3D dataset with anomaly-free training data and annotated defect test samples for unsupervised anomaly detection in industrial settings.

Findings

Initial benchmark shows significant room for improvement in 3D anomaly detection methods.

Dataset covers various defect types including scratches, dents, and deformations.

High-resolution 3D sensor data enables detailed defect localization.

Abstract

We introduce the first comprehensive 3D dataset for the task of unsupervised anomaly detection and localization. It is inspired by real-world visual inspection scenarios in which a model has to detect various types of defects on manufactured products, even if it is trained only on anomaly-free data. There are defects that manifest themselves as anomalies in the geometric structure of an object. These cause significant deviations in a 3D representation of the data. We employed a high-resolution industrial 3D sensor to acquire depth scans of 10 different object categories. For all object categories, we present a training and validation set, each of which solely consists of scans of anomaly-free samples. The corresponding test sets contain samples showing various defects such as scratches, dents, holes, contaminations, or deformations. Precise ground-truth annotations are provided for every anomalous test sample. An initial benchmark of 3D anomaly detection methods on our dataset indicates a considerable room for improvement.