GVGS: Gaussian Visibility-Aware Multi-View Geometry for Accurate Surface Reconstruction

TL;DR

This paper introduces GVGS, a novel multi-view geometry method that explicitly models Gaussian visibility to improve surface reconstruction accuracy, addressing limitations of depth-based visibility inference.

Contribution

It proposes a Gaussian visibility-aware multi-view geometric consistency formulation and a progressive depth calibration strategy to enhance 3D reconstruction.

Findings

Improves reconstruction accuracy on DTU and Tanks and Temples datasets.

Explicit Gaussian visibility modeling outperforms traditional depth-based methods.

Open-sourced code at https://github.com/GVGScode/GVGS.

Abstract

3D Gaussian Splatting (3DGS) enables efficient rendering, yet accurate surface reconstruction remains challenging due to unreliable geometric supervision. Existing approaches predominantly rely on depth-based reprojection to infer visibility and enforce multi-view consistency, leading to a fundamental circular dependency: visibility estimation requires accurate depth, while depth supervision itself is conditioned on visibility. In this work, we revisit multi-view geometric supervision from the perspective of visibility modeling. Instead of inferring visibility from pixel-wise depth consistency, we explicitly model visibility at the level of Gaussian primitives. We introduce a Gaussian visibility-aware multi-view geometric consistency (GVMV) formulation, which aggregates cross-view visibility of shared Gaussians to construct reliable supervision over co-visible regions. To further incorporate monocular priors, we propose a progressive quadtree-calibrated depth alignment (QDC) strategy that performs block-wise affine calibration under visibility-aware guidance, effectively mitigating scale ambiguity while preserving local geometric structures. Extensive experiments on DTU and Tanks and Temples demonstrate that our method consistently improves reconstruction accuracy over prior Gaussian-based approaches. Our code is fully open-sourced and available at an anonymous repository: https://github.com/GVGScode/GVGS.