Face recognizing script using OpenCV
Face recognition is one of the challenging tasks when it comes to models like
CNN - Convolutional Neural Networks
HOG - Histogram Oriented Gradients
SIFT - Scale Invariant Feature Transform
It all depends on how many images you train your model with and how good/bad is the acceptance/tolerance level.
For the type of images I used, I found the model 'hog' to be best with 0.6 learning rate. However these values can alter based on various factors.