Gaussian Head Avatar: Ultra High-fidelity Head Avatar via Dynamic
  Gaussians

Yuelang Xu; Benwang Chen; Zhe Li; Hongwen Zhang; Lizhen Wang; Zerong Zheng; Yebin Liu

返回文献探索

Paper · arXiv 2312.03029

Gaussian Head Avatar: Ultra High-fidelity Head Avatar via Dynamic Gaussians

Yuelang Xu, Benwang Chen, Zhe Li, Hongwen Zhang, Lizhen Wang, Zerong Zheng, Yebin Liu

27 upvotesDecember 5, 2023arXiv 预印本

AI 摘要

The method uses controllable 3D Gaussians and a fully learned MLP-based deformation field for high-fidelity 3D head avatar modeling under sparse views, employing geometry-guided initialization with implicit SDF and Deep Marching Tetrahedra for stability and high rendering quality.

3D GaussiansMLP-based deformation fieldimplicit SDFDeep Marching Tetrahedrahigh-fidelity 3D head avatarssparse viewsultra high-fidelity rendering

Abstract

Creating high-fidelity 3D head avatars has always been a research hotspot, but there remains a great challenge under lightweight sparse view setups. In this paper, we propose Gaussian Head Avatar represented by controllable 3D Gaussians for high-fidelity head avatar modeling. We optimize the neutral 3D Gaussians and a fully learned MLP-based deformation field to capture complex expressions. The two parts benefit each other, thereby our method can model fine-grained dynamic details while ensuring expression accuracy. Furthermore, we devise a well-designed geometry-guided initialization strategy based on implicit SDF and Deep Marching Tetrahedra for the stability and convergence of the training procedure. Experiments show our approach outperforms other state-of-the-art sparse-view methods, achieving ultra high-fidelity rendering quality at 2K resolution even under exaggerated expressions.