This is an unofficial implementation of the paper “Student-Teacher Feature Pyramid Matching for Unsupervised Anomaly Detection”.

STPM-Anomaly-Detection-Localization-master

This is an implementation of the paper Student-Teacher Feature Pyramid Matching for Unsupervised Anomaly Detection.

Datasets

MVTec AD datasets : Download from MVTec website

Environment

pip install -r requirements.txt

Usage

python main.py --phase 'train or test' --dataset_path 'D:/dataset/mvtec_anomaly_detection' --save_path 'path\to\save\results' --obj 'class name'

MVTecAD AUC-ROC score (mean of n trials)

Visualization examples

Acknowledgement

The code is partially adapted from STPM_anomaly_detection