V-Express aims to generate a talking head video under the control of a reference image, an audio, and a sequence of V-Kps images. - View it on GitHub