An Outlier Exposure Approach to Improve Visual Anomaly Detection Performance for Mobile Robots

TL;DR

This paper introduces an outlier exposure method to enhance visual anomaly detection in mobile robots, demonstrating that even limited anomalous data can significantly improve detection accuracy.

Contribution

The paper proposes a novel outlier exposure approach combined with Real-NVP for anomaly detection, specifically tailored for robotics applications with scarce anomalous data.

Findings

Outperforms existing methods on a new indoor patrolling dataset

Exposing few anomalous frames significantly improves detection performance

Method effectively leverages limited anomalous data in robotics scenarios

Abstract

We consider the problem of building visual anomaly detection systems for mobile robots. Standard anomaly detection models are trained using large datasets composed only of non-anomalous data. However, in robotics applications, it is often the case that (potentially very few) examples of anomalies are available. We tackle the problem of exploiting these data to improve the performance of a Real-NVP anomaly detection model, by minimizing, jointly with the Real-NVP loss, an auxiliary outlier exposure margin loss. We perform quantitative experiments on a novel dataset (which we publish as supplementary material) designed for anomaly detection in an indoor patrolling scenario. On a disjoint test set, our approach outperforms alternatives and shows that exposing even a small number of anomalous frames yields significant performance improvements.