Towards Zero-shot 3D Anomaly Localization

TL;DR

This paper introduces 3DzAL, a novel zero-shot 3D anomaly detection framework that leverages contrastive learning and pseudo anomalies to identify anomalies without prior class overlap, advancing industrial inspection capabilities.

Contribution

The paper proposes a new zero-shot 3D anomaly detection method using patch-level contrastive learning and adversarial perturbations, addressing data privacy and class mismatch issues.

Findings

3DzAL outperforms existing methods in anomaly detection accuracy.

The approach effectively handles unseen classes in 3D anomaly localization.

Contrastive learning with pseudo anomalies improves feature representation.

Abstract

3D anomaly detection and localization is of great significance for industrial inspection. Prior 3D anomaly detection and localization methods focus on the setting that the testing data share the same category as the training data which is normal. However, in real-world applications, the normal training data for the target 3D objects can be unavailable due to issues like data privacy or export control regulation. To tackle these challenges, we identify a new task -- zero-shot 3D anomaly detection and localization, where the training and testing classes do not overlap. To this end, we design 3DzAL, a novel patch-level contrastive learning framework based on pseudo anomalies generated using the inductive bias from task-irrelevant 3D xyz data to learn more representative feature representations. Furthermore, we train a normalcy classifier network to classify the normal patches and pseudo anomalies and utilize the classification result jointly with feature distance to design anomaly scores. Instead of directly using the patch point clouds, we introduce adversarial perturbations to the input patch xyz data before feeding into the 3D normalcy classifier for the classification-based anomaly score. We show that 3DzAL outperforms the state-of-the-art anomaly detection and localization performance.