Learning Discriminative Signed Distance Functions from Multi-scale Level-of-detail Features for 3D Anomaly Detection

2026-05-05

AI Summary

This work addresses the challenge of point-level anomaly detection in large-scale sparse 3D point clouds by proposing an implicit surface-based multi-scale representation learning approach. The method enhances anomaly awareness through a noise point generation module, captures both local and global geometric cues via hierarchical multi-scale feature extraction, and, for the first time, integrates signed distance functions (SDFs) with multi-scale features to construct an implicit surface discrimination module that effectively distinguishes normal from anomalous points. Evaluated on the Anomaly-ShapeNet and Real3D-AD datasets, the proposed method achieves average object-level AUROC scores of 92.1% and 85.9%, respectively, surpassing the current state-of-the-art by 2.1% and 3.6%, thereby significantly advancing the performance frontier in 3D point cloud anomaly detection.

Abstract

Detecting anomalies from 3D point clouds has received increasing attention in the field of computer vision, with some group-based or point-based methods achieving impressive results in recent years. However, learning accurate point-wise representations for 3D anomaly detection faces great challenges due to the large scale and sparsity of point clouds. In this study, a surface-based method is proposed for 3D anomaly detection, which learns a discriminative signed distance function using multi-scale level-of-detail features. We first present a Noisy Points Generation (NPG) module to generate different types of noise, thereby facilitating the learning of discriminative features by exposing abnormal points. Then, we introduce a Multi-scale Level-of-detail Feature (MLF) module to capture multi-scale information from a point cloud, which provides both fine-grained local and coarse-grained global feature information. Finally, we design an Implicit Surface Discrimination (ISD) module that leverages the extracted multi-scale features to learn an implicit surface representation of point clouds, which effectively trains a signed distance function to distinguish between abnormal and normal points. Experimental results demonstrate that the proposed method achieves an average object-level AUROC of 92.1\% and 85.9\% on the Anomaly-ShapeNet and Real3D-AD datasets, outperforming the current best approach by 2.1\% and 3.6\%, respectively. Codes are available at https://anonymous.4open.science/r/DLF-3AD-DA61.

Problem

• 3D anomaly detection
• point clouds
• signed distance function
• multi-scale features
• anomaly representation

Innovation

• Signed Distance Function
• Multi-scale Features
• Implicit Surface Representation
• 3D Anomaly Detection
• Noisy Points Generation