ReConPatch : Contrastive Patch Representation Learning for Industrial Anomaly Detection

Open paper
Benchmark Model Rank Results
anomaly-detection-on-btadReConPatch WRN-50#6Detection AUROC: 95.8Segmentation AUPRO: 97.5
anomaly-detection-on-mvtec-adReConPatch Ensemble (+RefineNet)#10Detection AUROC: 99.72Segmentation AUROC: 99.2
anomaly-detection-on-mvtec-adReConPatch WRN-50 (+RefineNet)#11Detection AUROC: 99.71Segmentation AUROC: 98.62
anomaly-detection-on-mvtec-adReConPatch WRN-101#18Detection AUROC: 99.62Segmentation AUROC: 98.53
anomaly-detection-on-mvtec-adReConPatch WRN-50#26Detection AUROC: 99.56Segmentation AUROC: 98.18
anomaly-detection-on-mvtec-adReConPatch Ensemble#116Segmentation AUROC: 98.67

Rank counts only results with a code link.