Training data(faces) are stored in a folder called 'train', with each file identified by the ID of the person and an index value.
First train the model on the faces and then use the webcam and opencv to label faces detected by the webcam.