ObamaNet: Photo-realistic lip-sync from text

TL;DR

ObamaNet is a fully trainable neural architecture that converts any text into synchronized audio and photo-realistic lip-sync videos without traditional graphics methods.

Contribution

It introduces a novel neural pipeline combining text-to-speech, keypoint generation, and video synthesis for realistic lip-sync video creation from text.

Findings

First fully trainable neural lip-sync system

Generates synchronized audio and video from text

Does not rely on traditional graphics techniques

Abstract

We present ObamaNet, the first architecture that generates both audio and synchronized photo-realistic lip-sync videos from any new text. Contrary to other published lip-sync approaches, ours is only composed of fully trainable neural modules and does not rely on any traditional computer graphics methods. More precisely, we use three main modules: a text-to-speech network based on Char2Wav, a time-delayed LSTM to generate mouth-keypoints synced to the audio, and a network based on Pix2Pix to generate the video frames conditioned on the keypoints.