3D Anomaly Detection on Real 3D-AD

Metric: Point AUROC (higher is better)

LeaderboardDataset

Loading chart...

Results

Hide extra data

Sort:

#	Model↕	Point AUROC▼	Extra Data	Paper	Date↕	Code
1	GLFM	0.898	No	-	-	Code
2	DUS-Net	0.847	No	Taming Anomalies with Down-Up Sampling Networks:...	2025-07-05	-
3	MC4AD	0.837	Yes	Examining the Source of Defects from a Mechanica...	2025-05-09	Code
4	ISMP	0.836	No	Look Inside for More: Internal Spatial Modality ...	2024-12-18	Code
5	MC3D-AD	0.768	No	MC3D-AD: A Unified Geometry-aware Reconstruction...	2025-05-04	-
6	PASDF	0.745	No	Bridging 3D Anomaly Localization and Repair via ...	2025-05-30	-
7	Group3AD	0.735	No	Towards High-resolution 3D Anomaly Detection via...	2024-08-08	-
8	BTF (Raw)	0.722	No	Back to the Feature: Classical 3D Features are (...	2022-03-10	Code
9	PointAD	0.716	No	PointAD: Comprehending 3D Anomalies from Points ...	2024-10-01	Code
10	Reg3D-AD	0.7	Yes	Real3D-AD: A Dataset of Point Cloud Anomaly Dete...	2023-09-23	Code
11	PatchCore (FPFH+Raw)	0.692	No	Towards Total Recall in Industrial Anomaly Detec...	2021-06-15	Code
12	M3DM (PointMAE)	0.637	No	Multimodal Industrial Anomaly Detection via Hybr...	2023-03-01	Code
13	M3DM (PointBERT)	0.636	No	Multimodal Industrial Anomaly Detection via Hybr...	2023-03-01	Code
14	PatchCore (PointMAE)	0.634	No	Towards Total Recall in Industrial Anomaly Detec...	2021-06-15	Code
15	R3D-AD	0.592	No	R3D-AD: Reconstruction via Diffusion for 3D Anom...	2024-07-15	Code
16	PatchCore (FPFH)	0.592	No	Towards Total Recall in Industrial Anomaly Detec...	2021-06-15	Code
17	BTF (FPFH)	0.566	No	Back to the Feature: Classical 3D Features are (...	2022-03-10	Code

#1GLFM
0.898
Point AUROC
No paperCode
#2DUS-NetSOTA
0.847
Point AUROC· 2025-07-05
Taming Anomalies with Down-Up Sampling Networks: Group Center Preserving Reconstruction for 3D Anomaly Detection
#3MC4ADSOTA
0.837
Point AUROC· Extra Data· 2025-05-09
Examining the Source of Defects from a Mechanical Perspective for 3D Anomaly Detection Code
#4ISMPSOTA
0.836
Point AUROC· 2024-12-18
Look Inside for More: Internal Spatial Modality Perception for 3D Anomaly Detection Code
#5MC3D-AD
0.768
Point AUROC· 2025-05-04
MC3D-AD: A Unified Geometry-aware Reconstruction Model for Multi-category 3D Anomaly Detection
#6PASDF
0.745
Point AUROC· 2025-05-30
Bridging 3D Anomaly Localization and Repair via High-Quality Continuous Geometric Representation
#7Group3ADSOTA
0.735
Point AUROC· 2024-08-08
Towards High-resolution 3D Anomaly Detection via Group-Level Feature Contrastive Learning
#8BTF (Raw)SOTA
0.722
Point AUROC· 2022-03-10
Back to the Feature: Classical 3D Features are (Almost) All You Need for 3D Anomaly Detection Code
#9PointAD
0.716
Point AUROC· 2024-10-01
PointAD: Comprehending 3D Anomalies from Points and Pixels for Zero-shot 3D Anomaly Detection Code
#10Reg3D-AD
0.7
Point AUROC· Extra Data· 2023-09-23
Real3D-AD: A Dataset of Point Cloud Anomaly Detection Code
#11PatchCore (FPFH+Raw)SOTA
0.692
Point AUROC· 2021-06-15
Towards Total Recall in Industrial Anomaly Detection Code
#12M3DM (PointMAE)
0.637
Point AUROC· 2023-03-01
Multimodal Industrial Anomaly Detection via Hybrid Fusion Code
#13M3DM (PointBERT)
0.636
Point AUROC· 2023-03-01
Multimodal Industrial Anomaly Detection via Hybrid Fusion Code
#14PatchCore (PointMAE)
0.634
Point AUROC· 2021-06-15
Towards Total Recall in Industrial Anomaly Detection Code
#15R3D-AD
0.592
Point AUROC· 2024-07-15
R3D-AD: Reconstruction via Diffusion for 3D Anomaly Detection Code
#16PatchCore (FPFH)
0.592
Point AUROC· 2021-06-15
Towards Total Recall in Industrial Anomaly Detection Code
#17BTF (FPFH)
0.566
Point AUROC· 2022-03-10
Back to the Feature: Classical 3D Features are (Almost) All You Need for 3D Anomaly Detection Code