This is a re-implementaion repo for the paper "Expressive Talking Head Video Encoding in StyleGan2 Latent-Space."

Install

Demo

re-enactment.

Video1	Video2
demo1.mp4	validate_video.mp4

single image puppet.

driving video1

Item	Image1	Image2
sync	validate_video.mp4	pic_07.mp4

driving video 2

Image3	Image4
pic_05.mp4	pic_03.mp4

code pipeline

find the index from a dataset.