image clustering according to the faces they contain

[Minyar2004]

Hello !

I've just stumbled upon facenet-pytorch. according to the documentation, this module can detect and recognize faces.

In my case, I'd like to classifying 20,000 images based on the faces they contain. I need to detect faces within a set of images (MTCNN), then identify and determine if they appear in the other images (InceptionResnetV1).

For example, on a set of 20,000 images: face 1 appears on images 2124, 9874 and 11587. face 2 appears on images 12, 6879, 19871 and 19987

Is there a way to do this efficiently with this module?

Thanks