GSurf: 3D Reconstruction via Signed Distance Fields with Direct Gaussian Supervision

TL;DR

GSurf introduces a novel end-to-end approach for 3D surface reconstruction that directly learns signed distance fields from Gaussian primitives, achieving faster speeds and high-quality results comparable to existing neural implicit methods.

Contribution

It proposes a new method to learn signed distance fields directly from Gaussian primitives, improving speed and reconstruction quality over prior Gaussian splatting techniques.

Findings

Faster training and rendering speeds compared to traditional 3D Gaussian splatting.

Produces high-fidelity 3D reconstructions with fewer holes and fragmented surfaces.

Achieves results comparable to neural implicit surface methods on benchmark datasets.

Abstract

Surface reconstruction from multi-view images is a core challenge in 3D vision. Recent studies have explored signed distance fields (SDF) within Neural Radiance Fields (NeRF) to achieve high-fidelity surface reconstructions. However, these approaches often suffer from slow training and rendering speeds compared to 3D Gaussian splatting (3DGS). Current state-of-the-art techniques attempt to fuse depth information to extract geometry from 3DGS, but frequently result in incomplete reconstructions and fragmented surfaces. In this paper, we introduce GSurf, a novel end-to-end method for learning a signed distance field directly from Gaussian primitives. The continuous and smooth nature of SDF addresses common issues in the 3DGS family, such as holes resulting from noisy or missing depth data. By using Gaussian splatting for rendering, GSurf avoids the redundant volume rendering typically required in other GS and SDF integrations. Consequently, GSurf achieves faster training and rendering speeds while delivering 3D reconstruction quality comparable to neural implicit surface methods, such as VolSDF and NeuS. Experimental results across various benchmark datasets demonstrate the effectiveness of our method in producing high-fidelity 3D reconstructions.