High-Fidelity 3D Gaussian Human Reconstruction via Region-Aware Initialization and Geometric Priors

TL;DR

This paper introduces a novel 3D Gaussian human reconstruction method that combines region-aware initialization and geometric priors, achieving high-fidelity, real-time results with detailed geometry from RGB images.

Contribution

It proposes a new framework using SMPL-X initialization, region-aware density, and multi-scale hash encoding for improved detail and efficiency in 3D human reconstruction.

Findings

Achieves superior reconstruction quality on benchmark datasets.

Preserves fine geometric details during complex motions.

Maintains real-time rendering speed.

Abstract

Real-time, high-fidelity 3D human reconstruction from RGB images is essential for interactive applications such as virtual reality and gaming, yet remains challenging due to the complex non-rigid deformations of dynamic human bodies. Although 3D Gaussian Splatting enables efficient rendering, existing methods struggle to capture fine geometric details and often produce artifacts such as fused fingers and over-smoothed faces. Moreover, conventional spatial-field-based dynamic modeling faces a trade-off between reconstruction fidelity and GPU memory consumption. To address these issues, we propose a novel 3D Gaussian human reconstruction framework that combines region-aware initialization with rich geometric priors. Specifically, we leverage the expressive SMPL-X model to initialize both 3D Gaussians and skinning weights, providing a robust geometric foundation for precise reconstruction. We further introduce a region-aware density initialization strategy and a geometry-aware multi-scale hash encoding module to improve local detail recovery while maintaining computational efficiency.Experiments on PeopleSnapshot and GalaBasketball show that our method achieves superior reconstruction quality and finer detail preservation under complex motions, while maintaining real-time rendering speed.