RFS-Energy-Anomaly-Detection-of-Defect

Implemention of the paper "Anomaly Detection of Defect using Energy of Point Pattern Features within Random Finite Set Framework"

https://arxiv.org/abs/2108.12159

Anomaly Detection of Defect using Energy of Point Pattern Features within Random Finite Set Framework Results and Comparisons

• AUC Performance on MVTec-AD 94.1 % AUC of the ROC curve -Comparison provided against DifferNet, 1-NN, OCSVM, DSEBM, GANomaly, and GeoTrans.

ROC Performance MVTec AD

#Few-shot Experimental Results for MVTec AD

• AUC=89.02 % for Ten Shot
• Comparison provided against DifferNet, DROCC, PatchSVDD,DeepSVDD,GeoTrans, and HTDGM.

MVTecAD AUC vs shots

Samples of D2-Net Features

The source code of the paper is placed under code