NPGA: Neural Parametric Gaussian Avatars

TL;DR

NPGA introduces a novel data-driven approach for creating high-fidelity, controllable human head avatars from multi-view videos, leveraging 3D Gaussian splatting and neural parametric head models for improved realism and animation.

Contribution

The paper presents NPGA, a new method combining 3D Gaussian splatting with neural parametric head models to generate detailed, controllable avatars from videos, outperforming previous state-of-the-art techniques.

Findings

Outperforms previous avatars on the NeRSemble dataset with 2.6 PSNR improvement.

Demonstrates accurate animation from monocular videos.

Uses Gaussian primitives with latent features for dynamic expressivity.

Abstract

The creation of high-fidelity, digital versions of human heads is an important stepping stone in the process of further integrating virtual components into our everyday lives. Constructing such avatars is a challenging research problem, due to a high demand for photo-realism and real-time rendering performance. In this work, we propose Neural Parametric Gaussian Avatars (NPGA), a data-driven approach to create high-fidelity, controllable avatars from multi-view video recordings. We build our method around 3D Gaussian splatting for its highly efficient rendering and to inherit the topological flexibility of point clouds. In contrast to previous work, we condition our avatars' dynamics on the rich expression space of neural parametric head models (NPHM), instead of mesh-based 3DMMs. To this end, we distill the backward deformation field of our underlying NPHM into forward deformations which are compatible with rasterization-based rendering. All remaining fine-scale, expression-dependent details are learned from the multi-view videos. For increased representational capacity of our avatars, we propose per-Gaussian latent features that condition each primitives dynamic behavior. To regularize this increased dynamic expressivity, we propose Laplacian terms on the latent features and predicted dynamics. We evaluate our method on the public NeRSemble dataset, demonstrating that NPGA significantly outperforms the previous state-of-the-art avatars on the self-reenactment task by 2.6 PSNR. Furthermore, we demonstrate accurate animation capabilities from real-world monocular videos.