This is an unofficial implementation of the paper “PaDiM: a Patch Distribution Modeling Framework for Anomaly Detection and Localization”.

This project proposes an end-to-end framework for semi-supervised Anomaly Detection and Segmentation in images based on Deep Learning.

Official PyTorch code for WACV 2022 paper "CFLOW-AD: Real-Time Unsupervised Anomaly Detection with Localization via Conditional Normalizing Flows"

Code underlying our publication "Modeling the Distribution of Normal Data in Pre-Trained Deep Features for Anomaly Detection" at ICPR2020

Uninformed Students: Student-Teacher Anomaly Detection with Discriminative Latent Embeddings

Reconstruction by Inpainting Based Anomaly Detection

Anomaly detection method that incorporates multi-scale features to sparse coding

WIP: Unofficial Tensorflow 2.x Implementation of ReConPatch (https://arxiv.org/abs/2305.16713)

PaDiM: a Patch Distribution Modeling Framework for Anomaly Detection and Localization

🪥 Unofficial re-implementation of Semi-orthogonal Embedding for Efficient Unsupervised Anomaly Segmentation

EfficientNetV2 based PaDiM

A Curated List of Awesome Unsupervised Anomaly Detection on MVTec AD Dataset

Thesis project about Visual Anomaly Detection based on Self Supervised Learning. The model identifies anomalies from information acquired during training, where normality and anomaly patterns are built using syntetic data

Grad-CAM implementation on MVTec dataset re-casted as a supervised learning problem.

This is an autoencoder implementation that was trained on MNIST and MVTEC Datasets to predict numbers and transistor position.

Implementation of paper: Rádli, R., & Czúni, L. (2021). About the Application of Autoencoders for Visual Defect Detection.

Cookiecutter Template for Unsupervised Anomaly Detection on MVTec AD Dataset

Same Same But DifferNet: Semi-Supervised Defect Detection with Normalizing Flows

Dockerfile for Unsupervised Anomaly Detection on MVTec AD Dataset