Multimodal Industrial Anomaly Detection via Hybrid Fusion
This addresses a specific problem in industrial quality control by improving anomaly detection accuracy, though it appears incremental as it builds on existing multimodal approaches.
The paper tackles the problem of multimodal industrial anomaly detection using 3D point clouds and RGB images, where existing methods suffer from feature disturbance due to direct concatenation. It proposes M3DM with hybrid fusion, achieving state-of-the-art performance on the MVTec-3D AD dataset in detection and segmentation precision.
2D-based Industrial Anomaly Detection has been widely discussed, however, multimodal industrial anomaly detection based on 3D point clouds and RGB images still has many untouched fields. Existing multimodal industrial anomaly detection methods directly concatenate the multimodal features, which leads to a strong disturbance between features and harms the detection performance. In this paper, we propose Multi-3D-Memory (M3DM), a novel multimodal anomaly detection method with hybrid fusion scheme: firstly, we design an unsupervised feature fusion with patch-wise contrastive learning to encourage the interaction of different modal features; secondly, we use a decision layer fusion with multiple memory banks to avoid loss of information and additional novelty classifiers to make the final decision. We further propose a point feature alignment operation to better align the point cloud and RGB features. Extensive experiments show that our multimodal industrial anomaly detection model outperforms the state-of-the-art (SOTA) methods on both detection and segmentation precision on MVTec-3D AD dataset. Code is available at https://github.com/nomewang/M3DM.