Thanks, great and simple tutorial! Seems to me it can be improved with more training examples and I might add that for production dlib's face detector is a bit slow. Also it would be nice to see how the system performs extracting facial features with dlib.shape_predictor.
You are viewing a single comment's thread from: