Anomaly detection with Deep learning method

Issue

This Content is from Stack Overflow. Question asked by tomtels

For anomaly detection using an image, several method based on DNN are proposed such as CNN, AutoEncoder, GAN, etc. I found this web page: https://paperswithcode.com/sota/anomaly-detection-on-mvtec-ad where proposed recent DNN methods are ranked using a particlar dataset.

According to the page, PatchCore (based on CNN ?) and FastFlow ( based on Transformer?) show the best performance for anomaly detection using dataset: MVTec-AD.

Huwever, the other methods based on AutoEncoder or GAN are not listed in the graph shown in a page of the link above. This means that AutoEncoder and GAN are not usefull for the anomaly detection?

Do some one know how much score AutoEncoder base or GAN base can achive for anomaly detection compared to PatchCore and FastFlow??



Solution

This question is not yet answered, be the first one who answer using the comment. Later the confirmed answer will be published as the solution.

This Question and Answer are collected from stackoverflow and tested by JTuto community, is licensed under the terms of CC BY-SA 2.5. - CC BY-SA 3.0. - CC BY-SA 4.0.

people found this article helpful. What about you?