Home/Browse/Industrial Inspection/Anomaly Detection/MVTec AD

MVTec AD

Name: MVTec AD Benchmark Results
Creator: Unknown
License: https://creativecommons.org/licenses/by/4.0/

Unknown

5,354 high-resolution images across 15 object and texture categories. The gold standard for industrial anomaly detection with pixel-level annotations.

Paper Leaderboard

Benchmark Stats

Models11

Papers11

Metrics2

SOTA History

Not enough data to show trend.

Image AUROC

Higher is better

Rank	Model	Source	Score	Year	Paper
1	SimpleNet CVPR 2023 SimpleNet, image-level AUROC across 15 categories	Editorial	99.6	2026	Source
2	PatchCore Memory-bank based anomaly detection, widely-cited baseline	Editorial	99.1	2026	Source
3	EfficientAD WACV 2024, real-time anomaly detection	Editorial	99.1	2026	Source

auroc

Higher is better

Rank	Model	Source	Score	Year	Paper
1	simplenet State-of-the-art on MVTec AD. Simple architecture.	Editorial	99.6	2025	Source
2	fastflow 2D normalizing flow. Good speed-accuracy tradeoff.	Editorial	99.4	2025	Source
3	patchcore Image-level AUROC. State-of-the-art with WideResNet-50 backbone.	Editorial	99.1	2025	Source
4	efficientad Matches PatchCore with 614 FPS inference.	Editorial	99.1	2025	Source
5	reverse-distillation Knowledge distillation approach.	Editorial	98.5	2025	Source
6	cflow-ad Conditional normalizing flows.	Editorial	98.3	2025	Source
7	draem Discriminatively trained reconstruction.	Editorial	98	2025	Source
8	padim ResNet-18 backbone. Fast inference.	Editorial	97.9	2025	Source

Submit a Result

Back to Anomaly Detection