Hi all,
I was given a project realating to Face Expression Recognition. The project just started a few day ago and there re something make me so confused. Based on some information from the google, i know that MHI is a technique to detect face for emotion classification by subtracting each frame from the first frame, but how can we know that the object subtracted is exactly a face. Some journal i read use a neural network to classify whether the object is a face or not but i think it 's a little bit complicated because most of the journals use a neural network for the last step: expression classification. That means if a system use MHI in the beginning, it have to use neural network twice. So, is there any suggestion for the MHI method? how to know the object after MHI method is really a face? Thanks
Sorry if this post is ridiculous, i am only the first year CS student and this is my first research in the university, so there re something i m not used to