Point Cloud Diffusion with Global and Local Reconstruction for Instance-Level 3D Anomaly Detection

2026-06-24

AI Summary

This work addresses the challenges of reconstructing and detecting subtle defects—such as scratches—in point clouds, as well as false positives caused by reconstruction biases in background regions. To this end, the authors propose PCDiff, a novel framework that introduces, for the first time, an instance-level multimodal conditional diffusion model to 3D anomaly generation. By integrating texture gradients, image patches, textual descriptions, and mask conditions, PCDiff enables high-fidelity modeling of weak anomalies during generation. For detection, it employs a local–global joint reconstruction algorithm that simultaneously preserves fine-grained local details and maintains overall geometric consistency. The method significantly improves anomaly detection accuracy, effectively suppresses background false positives, and achieves superior performance over existing approaches in both 3D anomaly generation fidelity and reconstruction quality.

Abstract

3D anomaly detection in point clouds is critical for high-precision industrial manufacturing. Reconstruction-based methods have laid a strong foundation by detecting 3D anomalies through comparisons between defective inputs and their reconstructed normal counterparts. However, existing methods still suffer from two challenges: 1) the foreground weak defective regions such as scratches are hard to reconstruct and detect, where the anomaly deviations in normalized point clouds can be as small as $10^{-3}$; 2) the background non-defective regions are prone to get positional bias in reconstruction, which leads to false positives. To address these challenges, we propose \textbf{PCDiff}, a point cloud diffusion framework for instance-level 3D anomaly generation and detection. In the generation phase, an instance-level multi-modal attention is embedded into the generation framework, where anomalies are conditioned with texture gradient, image patch, text and mask. The instance-level condition enables the high-quality generation of weak-defective anomalies. In the detection phase, a joint local-global reconstruction algorithm is introduced to ensure local anomaly restoration and global geometric consistency, which preserves background normal structure while restoring the foreground defect. Extensive experiments demonstrate that the proposed PCDiff significantly outperforms state-of-the-art methods in both 3D anomaly generation fidelity and reconstruction quality, leading to substantial improvements in anomaly detection accuracy.

Problem

• 3D anomaly detection
• point cloud
• weak-defective regions
• positional bias
• reconstruction-based methods

Innovation

• point cloud diffusion
• instance-level anomaly detection
• multi-modal attention
• local-global reconstruction
• 3D anomaly generation