GauHuman: Articulated Gaussian Splatting from Monocular Human Videos

TL;DR

GauHuman is a fast, real-time 3D human modeling framework using Gaussian Splatting that significantly reduces training and rendering times while maintaining high quality, suitable for dynamic human performance capture.

Contribution

It introduces GauHuman, a novel Gaussian Splatting-based 3D human modeling method with efficient training, pose refinement, and pruning techniques for high-quality, real-time rendering.

Findings

Achieves state-of-the-art performance on ZJU_Mocap and MonoCap datasets.

Enables 3D human modeling with approximately 13,000 Gaussians in real-time.

Training time reduced to 1-2 minutes with rendering speeds up to 189 FPS.

Abstract

We present, GauHuman, a 3D human model with Gaussian Splatting for both fast training (1 ~ 2 minutes) and real-time rendering (up to 189 FPS), compared with existing NeRF-based implicit representation modelling frameworks demanding hours of training and seconds of rendering per frame. Specifically, GauHuman encodes Gaussian Splatting in the canonical space and transforms 3D Gaussians from canonical space to posed space with linear blend skinning (LBS), in which effective pose and LBS refinement modules are designed to learn fine details of 3D humans under negligible computational cost. Moreover, to enable fast optimization of GauHuman, we initialize and prune 3D Gaussians with 3D human prior, while splitting/cloning via KL divergence guidance, along with a novel merge operation for further speeding up. Extensive experiments on ZJU_Mocap and MonoCap datasets demonstrate that GauHuman achieves state-of-the-art performance quantitatively and qualitatively with fast training and real-time rendering speed. Notably, without sacrificing rendering quality, GauHuman can fast model the 3D human performer with ~13k 3D Gaussians.