Face Recognition and Identification

Libraries/Packages used: cv2, numpy, os, pickle, PIL, cascades

It is an attempt to recognize faces and identify FRIENDS cast in any video or photograph.

The file faces-train.py can be modified acccordingly for better results.

Trial and error method can be used to check "conf" and a suitable value should be used to enhance performance.

The number of entries in the images folder can be increased in order to extend the identification to different people.

The model is not very accurate( which is pretty obvious as it does not use any of deep learning libraries or neural network ) but it performs pretty good over different test cases.

Make sure to use the images of approximately the same size while training, and then resize them.

A good resolution webcam is an add on!