Towards High-fidelity 3D Talking Avatar with Personalized Dynamic Texture
This work solves the problem of creating more realistic and personalized talking avatars for applications like virtual reality and entertainment, representing a novel advancement rather than an incremental improvement.
The paper tackles the problem of generating high-fidelity 3D talking avatars by addressing the lack of dynamic texture in prior methods, introducing a diffusion-based framework that simultaneously generates facial motions and dynamic textures from speech, resulting in outperforming prior arts in motion synthesis and producing realistic textures.
Significant progress has been made for speech-driven 3D face animation, but most works focus on learning the motion of mesh/geometry, ignoring the impact of dynamic texture. In this work, we reveal that dynamic texture plays a key role in rendering high-fidelity talking avatars, and introduce a high-resolution 4D dataset \textbf{TexTalk4D}, consisting of 100 minutes of audio-synced scan-level meshes with detailed 8K dynamic textures from 100 subjects. Based on the dataset, we explore the inherent correlation between motion and texture, and propose a diffusion-based framework \textbf{TexTalker} to simultaneously generate facial motions and dynamic textures from speech. Furthermore, we propose a novel pivot-based style injection strategy to capture the complicity of different texture and motion styles, which allows disentangled control. TexTalker, as the first method to generate audio-synced facial motion with dynamic texture, not only outperforms the prior arts in synthesising facial motions, but also produces realistic textures that are consistent with the underlying facial movements. Project page: https://xuanchenli.github.io/TexTalk/.