Parametric Implicit Face Representation for Audio-Driven Facial Reenactment

TL;DR

This paper introduces a novel parametric implicit face representation that combines interpretability and expressive power, enabling high-quality, controllable audio-driven facial reenactment with improved realism and identity preservation.

Contribution

The work proposes a new parametric implicit face representation parameterized by 3D face model parameters, bridging explicit and implicit methods, along with techniques for better encoding, synthesis, and generalization.

Findings

Produces more realistic talking head videos than previous methods.

Maintains greater fidelity to speaker identities and styles.

Demonstrates improved generalization and quality in experiments.

Abstract

Audio-driven facial reenactment is a crucial technique that has a range of applications in film-making, virtual avatars and video conferences. Existing works either employ explicit intermediate face representations (e.g., 2D facial landmarks or 3D face models) or implicit ones (e.g., Neural Radiance Fields), thus suffering from the trade-offs between interpretability and expressive power, hence between controllability and quality of the results. In this work, we break these trade-offs with our novel parametric implicit face representation and propose a novel audio-driven facial reenactment framework that is both controllable and can generate high-quality talking heads. Specifically, our parametric implicit representation parameterizes the implicit representation with interpretable parameters of 3D face models, thereby taking the best of both explicit and implicit methods. In addition, we propose several new techniques to improve the three components of our framework, including i) incorporating contextual information into the audio-to-expression parameters encoding; ii) using conditional image synthesis to parameterize the implicit representation and implementing it with an innovative tri-plane structure for efficient learning; iii) formulating facial reenactment as a conditional image inpainting problem and proposing a novel data augmentation technique to improve model generalizability. Extensive experiments demonstrate that our method can generate more realistic results than previous methods with greater fidelity to the identities and talking styles of speakers.