Research map / Point clouds / Point cloud classification
Bridging 3D Anomaly Localization and Repair via High-Quality Continuous Geometric Representation
Bozhong Zheng, Jinye Gan, Xiaohao Xu, Xintao Chen, Wenqiao Li, Xiaonan Huang, Na Ni, Yingna Wu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add RANSAC coarse alignment (ablation) | PASDF PAM · Anomaly-ShapeNet | O-AUROC: 0.711 -> 0.900 | Improved |
| add ICP refinement after RANSAC (ablation) | PASDF PAM · Anomaly-ShapeNet | P-AUROC: 0.836 -> 0.897 | Improved |
| add iterative optimization in PAM (ablation) | PASDF PAM · Anomaly-ShapeNet | O-AUROC: 0.871 -> 0.900 | Improved |
| add iterative optimization in PAM (ablation) | PASDF PAM · Anomaly-ShapeNet | P-AUROC: 0.884 -> 0.897 | Improved |
| add positional encoding (ablation) | PASDF · Anomaly-ShapeNet | P-AUROC: 0.783 -> 0.897 | Improved |
| add positional encoding (ablation) | PASDF · Anomaly-ShapeNet | O-AUROC: 0.887 -> 0.900 | Improved |
| add positional encoding (ablation) | PASDF repair · Real3D-AD (repair) | Chamfer distance: 0.0255 -> 0.0203 | Improved |
| add positional encoding (ablation) | PASDF repair · Anomaly-ShapeNet (repair) | Chamfer distance: 0.0575 -> 0.0445 | Improved |
| add positional encoding (ablation) | PASDF repair · Real3D-AD (repair) | EMD: 0.0133 -> 0.0110 | Improved |
| add PAM pose alignment (ablation) | BTF(FPFH) · Anomaly-ShapeNet | O-AUROC: 0.528 -> 0.579 | Improved |
| add PAM pose alignment (ablation) | PatchCore(FPFH) · Anomaly-ShapeNet | O-AUROC: 0.568 -> 0.814 | Improved |
| add PAM pose alignment (ablation) | PatchCore(FPFH) · Anomaly-ShapeNet | P-AUROC: 0.580 -> 0.867 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- R3D-AD — R3D-AD: Reconstruction via Diffusion for 3D Anomaly Detection (ECCV 2024) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Towards Scalable 3D Anomaly Detection and Localization: A Benchmark via 3D Anomaly Synthesis and A Self-Supervised Learning Network (CVPR 2024 · similarity 0.89)
- Registration is a Powerful Rotation-Invariance Learner for 3D Anomaly Detection (NeurIPS 2025 · similarity 0.87)
- PO3AD: Predicting Point Offsets toward Better 3D Point Cloud Anomaly Detection (CVPR 2025 · similarity 0.86)
- PointAD: Comprehending 3D Anomalies from Points and Pixels for Zero-shot 3D Anomaly Detection (NeurIPS 2024 · similarity 0.85)
- Towards Generalizable 3D Anomaly Detection via Relational Inconsistency Modeling (NeurIPS 2026 · similarity 0.85)
- Align3D-AD: Cross-Modal Feature Alignment and Dual-Prompt Learning for Zero-shot 3D Anomaly Detection (NeurIPS 2026 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
