Window

I’ve captured all the scenes of “Andrew” from the movie Bicentennial Man and ran a facial landmark detection algorithm on the images. I’ve also wrote a software that is capable of doing facial landmark detection in real-time from webcam images. I’ve then created a Machine learning model (image-to-image translation) using conditional adversarial networks.

When you look at the camera, your face with all your facial expressions is translated into a robot with perceptible differences from your face; i.e., two people a man and a woman will be translated into a man and a woman robot. No two robots look the same. CNNs, Facial Landmark Recognition, Model Training, Real Time Image-to-Image Translation