Anomaly Detection On Mvtec Ad

Metrics

Detection AUROC

Results

Performance results of various models on this benchmark

		Paper Title	Repository
GLASS	99.9	A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial Anomaly Detection and Localization
DDAD	99.8	Anomaly Detection with Conditioned Denoising Diffusion Models
PBAS	99.8	Progressive Boundary Guided Anomaly Synthesis for Industrial Anomaly Detection
INP-Fomer ViT-L (model-unified multi-class)	99.8	Exploring Intrinsic Normal Prototypes within a Single Image for Universal Anomaly Detection
EfficientAD (early stopping)	99.8	EfficientAD: Accurate Visual Anomaly Detection at Millisecond-Level Latencies
Dinomaly ViT-L (model-unified multi-class)	99.77	Dinomaly: The Less Is More Philosophy in Multi-Class Unsupervised Anomaly Detection
ReConPatch Ensemble (+RefineNet)	99.72	ReConPatch : Contrastive Patch Representation Learning for Industrial Anomaly Detection
ReConPatch WRN-50 (+RefineNet)	99.71	ReConPatch : Contrastive Patch Representation Learning for Industrial Anomaly Detection
ADClick	99.7	Towards Efficient Pixel Labeling for Industrial Anomaly Detection and Localization	-
CPR-fast	99.7	Target before Shooting: Accurate Anomaly Detection and Localization under One Millisecond via Cascade Patch Retrieval
MSFlow	99.7	MSFlow: Multi-Scale Flow-based Framework for Unsupervised Anomaly Detection
CPR	99.7	Target before Shooting: Accurate Anomaly Detection and Localization under One Millisecond via Cascade Patch Retrieval
PNI Ensemble	99.63	PNI : Industrial Anomaly Detection using Position and Neighborhood Information
ReConPatch WRN-101	99.62	ReConPatch : Contrastive Patch Representation Learning for Industrial Anomaly Detection
PatchCore Large	99.6	Towards Total Recall in Industrial Anomaly Detection
SimpleNet	99.6	SimpleNet: A Simple Network for Image Anomaly Detection and Localization
WeakREST-Un	99.6	Industrial Anomaly Detection and Localization Using Weakly-Supervised Residual Transformers	-
RealNet	99.6	RealNet: A Feature Selection Network with Realistic Synthetic Anomaly for Anomaly Detection
Dinomaly ViT-B (model-unified multi-class)	99.60	Dinomaly: The Less Is More Philosophy in Multi-Class Unsupervised Anomaly Detection
RememberingNormality	99.6	Remembering Normality: Memory-guided Knowledge Distillation for Unsupervised Anomaly Detection	-

0 of 144 row(s) selected.

Command Palette

Anomaly Detection On Mvtec Ad

Metrics

Results