Anomaly Detection on MVTec AD

Metric: Segmentation AP (higher is better)

LeaderboardDataset

Loading chart...

Results

Hide augmentations

Sort:

#	Model↕	Segmentation AP▼	Augmentations	Paper	Date↕	Code
1	WeakREST-Block	87.6	Yes	Industrial Anomaly Detection and Localization Us...	2023-06-06	-
2	ADClick	86.4	Yes	Towards Efficient Pixel Labeling for Industrial ...	2024-07-03	-
3	CPR	86	No	Target before Shooting: Accurate Anomaly Detecti...	2023-08-13	Code
4	WeakREST-Un	83	No	Industrial Anomaly Detection and Localization Us...	2023-06-06	-
5	ADClick	82.9	Yes	Towards Efficient Pixel Labeling for Industrial ...	2024-07-03	-
6	CPR	82.7	Yes	Target before Shooting: Accurate Anomaly Detecti...	2023-08-13	Code
7	CPR-fast	82.3	Yes	Target before Shooting: Accurate Anomaly Detecti...	2023-08-13	Code
8	CPR-faster	80.6	Yes	Target before Shooting: Accurate Anomaly Detecti...	2023-08-13	Code
9	PRN	78.6	Yes	Prototypical Residual Networks for Anomaly Detec...	2022-12-05	-
10	DeSTSeg	75.8	Yes	DeSTSeg: Segmentation Guided Denoising Student-T...	2022-11-21	Code
11	TASAD	74.6	No	-	-	Code
12	AD-CLSCNFs	74.32	No	-	-	Code
13	URD	72.4	Yes	Unlocking the Potential of Reverse Distillation ...	2024-12-10	Code
14	INP-Fomer ViT-L (model-unified multi-class)	72.1	No	Exploring Intrinsic Normal Prototypes within a S...	2025-03-04	Code
15	GLAD	70.9	No	GLAD: Towards Better Reconstruction with Global ...	2024-06-11	Code
16	Dinomaly ViT-L (model-unified multi-class)	70.53	Yes	Dinomaly: The Less Is More Philosophy in Multi-C...	2024-05-23	Code
17	DSR	70.2	Yes	DSR -- A dual subspace re-projection network for...	2022-08-02	Code
18	DRAEM+SSPCAB	69.9	Yes	Self-Supervised Predictive Convolutional Attenti...	2021-11-17	Code
19	Dinomaly ViT-B (model-unified multi-class)	69.29	Yes	Dinomaly: The Less Is More Philosophy in Multi-C...	2024-05-23	Code
20	DRAEM	68.4	Yes	DRAEM -- A discriminatively trained reconstructi...	2021-08-17	Code
21	EfficientAD-S	65.9	No	EfficientAD: Accurate Visual Anomaly Detection a...	2023-03-25	Code
22	MuSc (zero-shot)	62.7	Yes	MuSc: Zero-Shot Industrial Anomaly Classificatio...	2024-01-30	Code
23	VCP-CLIP	49.4	No	VCP-CLIP: A visual context prompting model for z...	2024-07-17	Code
24	Student–Teacher AD (Multiscale)	45.5	No	Uninformed Students: Student-Teacher Anomaly Det...	2019-11-06	Code
25	APRIL-GAN（zero-shot）	40.8	No	APRIL-GAN: A Zero-/Few-Shot Anomaly Classificati...	2023-05-27	Code
26	ACR (zero-shot)	38.9	No	Zero-Shot Anomaly Detection via Batch Normalizat...	2023-02-15	Code

#1WeakREST-BlockSOTA
87.6
Segmentation AP· Augmentations· 2023-06-06
Industrial Anomaly Detection and Localization Using Weakly-Supervised Residual Transformers
#2ADClick
86.4
Segmentation AP· Augmentations· 2024-07-03
Towards Efficient Pixel Labeling for Industrial Anomaly Detection and Localization
#3CPR
86
Segmentation AP· 2023-08-13
Target before Shooting: Accurate Anomaly Detection and Localization under One Millisecond via Cascade Patch Retrieval Code
#4WeakREST-Un
83
Segmentation AP· 2023-06-06
Industrial Anomaly Detection and Localization Using Weakly-Supervised Residual Transformers
#5ADClick
82.9
Segmentation AP· Augmentations· 2024-07-03
Towards Efficient Pixel Labeling for Industrial Anomaly Detection and Localization
#6CPR
82.7
Segmentation AP· Augmentations· 2023-08-13
Target before Shooting: Accurate Anomaly Detection and Localization under One Millisecond via Cascade Patch Retrieval Code
#7CPR-fast
82.3
Segmentation AP· Augmentations· 2023-08-13
Target before Shooting: Accurate Anomaly Detection and Localization under One Millisecond via Cascade Patch Retrieval Code
#8CPR-faster
80.6
Segmentation AP· Augmentations· 2023-08-13
Target before Shooting: Accurate Anomaly Detection and Localization under One Millisecond via Cascade Patch Retrieval Code
#9PRNSOTA
78.6
Segmentation AP· Augmentations· 2022-12-05
Prototypical Residual Networks for Anomaly Detection and Localization
#10DeSTSegSOTA
75.8
Segmentation AP· Augmentations· 2022-11-21
DeSTSeg: Segmentation Guided Denoising Student-Teacher for Anomaly Detection Code
#11TASAD
74.6
Segmentation AP
No paperCode
#12AD-CLSCNFs
74.32
Segmentation AP
No paperCode
#13URD
72.4
Segmentation AP· Augmentations· 2024-12-10
Unlocking the Potential of Reverse Distillation for Anomaly Detection Code
#14INP-Fomer ViT-L (model-unified multi-class)
72.1
Segmentation AP· 2025-03-04
Exploring Intrinsic Normal Prototypes within a Single Image for Universal Anomaly Detection Code
#15GLAD
70.9
Segmentation AP· 2024-06-11
GLAD: Towards Better Reconstruction with Global and Local Adaptive Diffusion Models for Unsupervised Anomaly Detection Code
#16Dinomaly ViT-L (model-unified multi-class)
70.53
Segmentation AP· Augmentations· 2024-05-23
Dinomaly: The Less Is More Philosophy in Multi-Class Unsupervised Anomaly Detection Code
#17DSRSOTA
70.2
Segmentation AP· Augmentations· 2022-08-02
DSR -- A dual subspace re-projection network for surface anomaly detection Code
#18DRAEM+SSPCABSOTA
69.9
Segmentation AP· Augmentations· 2021-11-17
Self-Supervised Predictive Convolutional Attentive Block for Anomaly Detection Code
#19Dinomaly ViT-B (model-unified multi-class)
69.29
Segmentation AP· Augmentations· 2024-05-23
Dinomaly: The Less Is More Philosophy in Multi-Class Unsupervised Anomaly Detection Code
#20DRAEMSOTA
68.4
Segmentation AP· Augmentations· 2021-08-17
DRAEM -- A discriminatively trained reconstruction embedding for surface anomaly detection Code
#21EfficientAD-S
65.9
Segmentation AP· 2023-03-25
EfficientAD: Accurate Visual Anomaly Detection at Millisecond-Level Latencies Code
#22MuSc (zero-shot)
62.7
Segmentation AP· Augmentations· 2024-01-30
MuSc: Zero-Shot Industrial Anomaly Classification and Segmentation with Mutual Scoring of the Unlabeled Images Code
#23VCP-CLIP
49.4
Segmentation AP· 2024-07-17
VCP-CLIP: A visual context prompting model for zero-shot anomaly segmentation Code
#24Student–Teacher AD (Multiscale)SOTA
45.5
Segmentation AP· 2019-11-06
Uninformed Students: Student-Teacher Anomaly Detection with Discriminative Latent Embeddings Code
#25APRIL-GAN（zero-shot）
40.8
Segmentation AP· 2023-05-27
APRIL-GAN: A Zero-/Few-Shot Anomaly Classification and Segmentation Method for CVPR 2023 VAND Workshop Challenge Tracks 1&2: 1st Place on Zero-shot AD and 4th Place on Few-shot AD Code
#26ACR (zero-shot)
38.9
Segmentation AP· 2023-02-15
Zero-Shot Anomaly Detection via Batch Normalization Code