CNN in Dlib Face Detected

Question

I want to understand how cnn works dlib face detected , as far as I know the cnn parameters and everything are in the mmod_human_face_detector.dat file, is there any way for me to read it? Or is there any article about it for me to look into?

cnn_face_detector = dlib.cnn_face_detection_model_v1('mmod_human_face_detector.dat')

I searched github and dlib's articles but couldn't find it

Answer 1

The question is not quite clearly worded, but by the line of code you inserted I assume you are interested in a specific model rather than CNN in general.

The model listed by you is based on detecting objects with maximum margin, and you can find the article here: MMOD.