One-Shot Free-View Neural Talking-Head Synthesis for Video Conferencing

Ting-Chun Wang   Arun Mallya   Ming-Yu Liu  

NVIDIA Corporation

CVPR 2021 (Oral)

[Paper] [arXiv] [Demo] [Video] [Talk]

Abstract

We propose a neural talking-head video synthesis model and demonstrate its application to video conferencing. Our model learns to synthesize a talking-head video using a source image containing the target person's appearance and a driving video that dictates the motion in the output. Our motion is encoded based on a novel keypoint representation, where the identity-specific and motion-related information is decomposed unsupervisedly. Extensive experimental validation shows that our model outperforms competing methods on benchmark datasets. Moreover, our compact keypoint representation enables a video conferencing system that achieves the same visual quality as the commercial H.264 standard while only using one-tenth of the bandwidth. Besides, we show our keypoint representation allows the user to rotate the head during synthesis, which is useful for simulating a face-to-face video conferencing experience.

paper thumbnail

Paper

arXiv

Citation

Ting-Chun Wang, Arun Mallya, Ming-Yu Liu. "One-Shot Free-View Neural Talking-Head Synthesis for Video Conferencing."
In CVPR, 2021. Bibtex


Code

This work is based upon Imaginaire. For more details, please consult research inquiries.

Demo

Please visit this page for an online demo.

Our Example Results

 

Video Reconstruction

 

Head Rotation

 

Face Frontalization

 

Motion Transfer

 

Citation

If you find this useful for your research, please use the following.
@inproceedings{wang2021facevid2vid,
  title={One-Shot Free-View Neural Talking-Head Synthesis for Video Conferencing},
  author={Ting-Chun Wang and Arun Mallya and Ming-Yu Liu},
  booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
  year={2021}
}