PointCore: Efficient Unsupervised Point Cloud Anomaly Detector Using Local-Global Features

TL;DR

PointCore introduces an efficient unsupervised point cloud anomaly detection method that uses a single memory bank for local and global features, reducing computational costs while maintaining high detection accuracy.

Contribution

It proposes a novel framework that combines local-global features with a normalization ranking method, improving efficiency and robustness over existing multi-memory bank approaches.

Findings

Achieves the best detection and localization performance on Real3D-AD dataset.

Demonstrates competitive inference time compared to state-of-the-art methods.

Effectively handles outliers with the normalization ranking method.

Abstract

Three-dimensional point cloud anomaly detection that aims to detect anomaly data points from a training set serves as the foundation for a variety of applications, including industrial inspection and autonomous driving. However, existing point cloud anomaly detection methods often incorporate multiple feature memory banks to fully preserve local and global representations, which comes at the high cost of computational complexity and mismatches between features. To address that, we propose an unsupervised point cloud anomaly detection framework based on joint local-global features, termed PointCore. To be specific, PointCore only requires a single memory bank to store local (coordinate) and global (PointMAE) representations and different priorities are assigned to these local-global features, thereby reducing the computational cost and mismatching disturbance in inference. Furthermore, to robust against the outliers, a normalization ranking method is introduced to not only adjust values of different scales to a notionally common scale, but also transform densely-distributed data into a uniform distribution. Extensive experiments on Real3D-AD dataset demonstrate that PointCore achieves competitive inference time and the best performance in both detection and localization as compared to the state-of-the-art Reg3D-AD approach and several competitors.