3D Anomaly Detection And Segmentation On
评估指标
Detection AUROC
Segmentation AUPRO
评测结果
各个模型在此基准测试上的表现结果
| Paper Title | Repository | |||
|---|---|---|---|---|
| Shape-Guided (only SDF) | 0.916 | 0.931 | Shape-Guided: Shape-Guided Dual-Memory Learning for 3D Anomaly Detection | - |
| CPMF (2D+3D) | 0.9515 | 0.9293 | Complementary Pseudo Multimodal Feature for Point Cloud Anomaly Detection | |
| Back to the Feature: Classical 3D Features are (Almost) All You Need for 3D Anomaly Detection (FPFH) | 0.782 | 0.924 | Back to the Feature: Classical 3D Features are (Almost) All You Need for 3D Anomaly Detection | |
| CPMF (3D) | 0.8304 | 0.9230 | Complementary Pseudo Multimodal Feature for Point Cloud Anomaly Detection | |
| CPMF (2D) | 0.8918 | 0.9145 | Complementary Pseudo Multimodal Feature for Point Cloud Anomaly Detection | |
| 3D-ST_128 | - | 0.833 | Anomaly Detection in 3D Point Clouds using Deep Geometric Descriptors | - |
| Voxel GAN | 0.537 | 0.583 | The MVTec 3D-AD Dataset for Unsupervised 3D Anomaly Detection and Localization | |
| Voxel VM | 0.571 | 0.492 | The MVTec 3D-AD Dataset for Unsupervised 3D Anomaly Detection and Localization | |
| Voxel AE | 0.699 | 0.348 | The MVTec 3D-AD Dataset for Unsupervised 3D Anomaly Detection and Localization |
0 of 9 row(s) selected.