Pytorch implementation of paper "One-Shot Free-View Neural Talking-Head Synthesis for Video Conferencing"