run the facedatset.py file to generate your image dataset. once it's running, a green bounding box will be displayed on you image and this will show that it's capturing images. You can make faces or pretend that you are on an interview so that there can be variety in your images. You can repeat this step for many people so that you can have a bigger dataset. press 'q' to quit the program.

once you're satisfied with the number of images you've got. run the face_train.py file to generate a trainner.yml file. This file contains an opencv-matrix for data about different individuals in you training data.

once done, you can run the final product using the faces.py file.

check the cascade folder and experiment with the cascades.