cnn network predict face landmarks (68 points) and head pose (3d pose, yaw,roll,pitch). - View it on GitHub