Multi-Perspective Anomaly Detection

TL;DR

This paper introduces a multi-perspective anomaly detection method using deep support vector data description, employing fusion techniques and data augmentation, validated on a new dices dataset and benchmarked against MNIST, outperforming single-perspective methods.

Contribution

It presents the first approach to multi-perspective image anomaly detection combining fusion techniques with a new dataset and improved performance over existing single-perspective methods.

Findings

Achieved ROC AUC of 80% with augmentation and denoising.

Proposed multi-perspective approach outperforms single-perspective methods.

Introduced the dices dataset with over 2000 images and rare anomalies.

Abstract

Anomaly detection is a critical problem in the manufacturing industry. In many applications, images of objects to be analyzed are captured from multiple perspectives which can be exploited to improve the robustness of anomaly detection. In this work, we build upon the deep support vector data description algorithm and address multi-perspective anomaly detection using three different fusion techniques, i.e., early fusion, late fusion, and late fusion with multiple decoders. We employ different augmentation techniques with a denoising process to deal with scarce one-class data, which further improves the performance (ROC AUC $= 80\%$). Furthermore, we introduce the dices dataset, which consists of over 2000 grayscale images of falling dices from multiple perspectives, with 5\% of the images containing rare anomalies (e.g., drill holes, sawing, or scratches). We evaluate our approach on the new dices dataset using images from two different perspectives and also benchmark on the standard MNIST dataset. Extensive experiments demonstrate that our proposed {multi-perspective} approach exceeds the state-of-the-art {single-perspective anomaly detection on both the MNIST and dices datasets}. To the best of our knowledge, this is the first work that focuses on addressing multi-perspective anomaly detection in images by jointly using different perspectives together with one single objective function for anomaly detection.