Speech2Video Synthesis with 3D Skeleton Regularization and Expressive Body Poses

TL;DR

This paper introduces a novel speech-to-video synthesis method that combines 3D skeleton regularization, expressive body poses, and part attention mechanisms in a GAN to produce realistic, synchronized talking videos with rich body dynamics.

Contribution

The paper presents a new framework integrating 3D skeleton regularization, gesture dictionaries, and part attention in GANs for improved speech-driven video synthesis.

Findings

Outperforms previous state-of-the-art methods in user studies

Generates realistic and expressive body movements from speech

Efficiently learns meaningful gestures with limited data

Abstract

In this paper, we propose a novel approach to convert given speech audio to a photo-realistic speaking video of a specific person, where the output video has synchronized, realistic, and expressive rich body dynamics. We achieve this by first generating 3D skeleton movements from the audio sequence using a recurrent neural network (RNN), and then synthesizing the output video via a conditional generative adversarial network (GAN). To make the skeleton movement realistic and expressive, we embed the knowledge of an articulated 3D human skeleton and a learned dictionary of personal speech iconic gestures into the generation process in both learning and testing pipelines. The former prevents the generation of unreasonable body distortion, while the later helps our model quickly learn meaningful body movement through a few recorded videos. To produce photo-realistic and high-resolution video with motion details, we propose to insert part attention mechanisms in the conditional GAN, where each detailed part, e.g. head and hand, is automatically zoomed in to have their own discriminators. To validate our approach, we collect a dataset with 20 high-quality videos from 1 male and 1 female model reading various documents under different topics. Compared with previous SoTA pipelines handling similar tasks, our approach achieves better results by a user study.