Research map / Point clouds / Point cloud classification
Real3D-AD: A Dataset of Point Cloud Anomaly Detection
Jiaqi Liu, Guoyang Xie, ruitao chen, Xinpeng Li, Jinbao Wang, Yong Liu, Chengjie Wang, Feng Zheng
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 |
|---|---|---|---|
| use Reg3D-AD instead of the strongest baseline | Real3D-AD · Real3D-AD | object-level AUROC: 0.682 -> 0.704 | Improved |
| use Reg3D-AD instead of the strongest baseline | Real3D-AD · Real3D-AD | object-level AUPR: 0.676 -> 0.723 | Improved |
| use Reg3D-AD instead of the strongest baseline | Real3D-AD · Real3D-AD | point-level AUROC: 0.722 -> 0.700 | Hurt |
| use Reg3D-AD instead of the strongest baseline | Real3D-AD · Real3D-AD | point-level AUPR: 0.129 -> 0.113 | Hurt |
| reduce PointMAE point groups from 16384 to 8192 (ablation) | Reg3D-AD · Real3D-AD | object-level AUROC: 0.705 -> 0.666 | Hurt |
| reduce PointMAE point groups from 16384 to 8192 (ablation) | Reg3D-AD · Real3D-AD | point-level AUROC: 0.700 -> 0.681 | Hurt |
| reduce PointMAE point groups from 16384 to 8192 (ablation) | Reg3D-AD · Real3D-AD | object-level AUPR: 0.723 -> 0.653 | Hurt |
| reduce PointMAE point groups from 16384 to 8192 (ablation) | Reg3D-AD · Real3D-AD | point-level AUPR: 0.113 -> 0.120 | Improved |
| increase memory bank size from 10000 to 20000 (ablation) | Reg3D-AD · Real3D-AD | object-level AUROC: 0.705 -> 0.674 | Hurt |
| increase memory bank size from 10000 to 20000 (ablation) | Reg3D-AD · Real3D-AD | point-level AUROC: 0.700 -> 0.697 | No clear effect |
| increase memory bank size from 10000 to 20000 (ablation) | Reg3D-AD · Real3D-AD | object-level AUPR: 0.723 -> 0.677 | Hurt |
| increase memory bank size from 10000 to 20000 (ablation) | Reg3D-AD · Real3D-AD | point-level AUPR: 0.113 -> 0.108 | Hurt |
Closest papers by meaning
- Registration is a Powerful Rotation-Invariance Learner for 3D Anomaly Detection (NeurIPS 2025 · similarity 0.91)
- Towards Scalable 3D Anomaly Detection and Localization: A Benchmark via 3D Anomaly Synthesis and A Self-Supervised Learning Network (CVPR 2024 · similarity 0.90)
- PO3AD: Predicting Point Offsets toward Better 3D Point Cloud Anomaly Detection (CVPR 2025 · similarity 0.90)
- Real-IAD D3: A Real-World 2D/Pseudo-3D/3D Dataset for Industrial Anomaly Detection (CVPR 2025 · similarity 0.89)
- Towards Generalizable 3D Anomaly Detection via Relational Inconsistency Modeling (NeurIPS 2026 · similarity 0.89)
- 3CAD: A Large-Scale Real-World 3C Product Dataset for Unsupervised Anomaly Detection (AAAI 2025 · similarity 0.88)
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.
