GeoSplat: A Deep Dive into Geometry-Constrained Gaussian Splatting

• URL: http://arxiv.org/abs/2509.05075v3
• Date: Fri, 26 Sep 2025 18:09:08 GMT
• Title: GeoSplat: A Deep Dive into Geometry-Constrained Gaussian Splatting
• Authors: Yangming Li, Chaoyu Liu, Lihao Liu, Simon Masnou, Carola-Bibiane Schönlieb,
• Abstract summary: We first present GeoSplat, a general geometry-constrained optimization framework that exploits both first-order and second-order geometric quantities.<n>We then introduce efficient and noise-robust estimation methods that provide dynamic geometric priors for our framework.<n>We conduct extensive experiments on multiple datasets for novel view synthesis, showing that our framework, GeoSplat, significantly improves the performance of Gaussian splatting.
• Score: 29.756579835391516
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: A few recent works explored incorporating geometric priors to regularize the optimization of Gaussian splatting, further improving its performance. However, those early studies mainly focused on the use of low-order geometric priors (e.g., normal vector), and they might also be unreliably estimated by noise-sensitive methods, like local principal component analysis. To address their limitations, we first present GeoSplat, a general geometry-constrained optimization framework that exploits both first-order and second-order geometric quantities to improve the entire training pipeline of Gaussian splatting, including Gaussian initialization, gradient update, and densification. As an example, we initialize the scales of 3D Gaussian primitives in terms of principal curvatures, leading to a better coverage of the object surface than random initialization. Secondly, based on certain geometric structures (e.g., local manifold), we introduce efficient and noise-robust estimation methods that provide dynamic geometric priors for our framework. We conduct extensive experiments on multiple datasets for novel view synthesis, showing that our framework, GeoSplat, significantly improves the performance of Gaussian splatting and outperforms previous baselines.

Related papers

• Geometry-Grounded Gaussian Splatting [46.209203640694994]
  We present a theoretical framework that establishes Gaussian primitives as a specific derivation type of solids.<n>We show that our method achieves the best shape reconstruction results among all Gaussian Splatting-based methods on public datasets.
  arXiv  Detail & Related papers  (2026-01-25T13:32:53Z)
• A Mixed-Primitive-based Gaussian Splatting Method for Surface Reconstruction [61.205927223522174]
  We propose a novel framework that enables Gaussian Splatting to incorporate multiple types of primitives during its surface reconstruction process.<n>Specifically, in our framework, we first propose a compositional splatting strategy, enabling the splatting and rendering of different types of primitives.
  arXiv  Detail & Related papers  (2025-07-15T13:52:40Z)
• QuickSplat: Fast 3D Surface Reconstruction via Learned Gaussian Initialization [69.50126552763157]
  Surface reconstruction is fundamental to computer vision and graphics, enabling applications in 3D modeling, mixed reality, robotics, and more.<n>Existing approaches based on rendering obtain promising results, but optimize on a per-scene basis, resulting in a slow optimization that can struggle to model textureless regions.<n>We introduce QuickSplat, which learns data-driven priors to generate dense initializations for 2D gaussian splatting optimization of large-scale indoor scenes.
  arXiv  Detail & Related papers  (2025-05-08T18:43:26Z)
• Sparse2DGS: Geometry-Prioritized Gaussian Splatting for Surface Reconstruction from Sparse Views [45.125032766506536]
  We propose Sparse2DGS, an MVS-d Gaussian Splatting pipeline for complete and accurate reconstruction.<n>Our key insight is to incorporate the geometric-prioritized enhancement schemes, allowing for direct and robust geometric learning under ill-posed conditions.<n>Sparse2DGS outperforms existing methods by notable margins while being $2times$ faster than the NeRF-based fine-tuning approach.
  arXiv  Detail & Related papers  (2025-04-29T02:47:02Z)
• FeatureGS: Eigenvalue-Feature Optimization in 3D Gaussian Splatting for Geometrically Accurate and Artifact-Reduced Reconstruction [1.474723404975345]
  3D Gaussian Splatting (3DGS) has emerged as a powerful approach for 3D scene reconstruction using 3D Gaussians.<n>We present FeatureGS, which incorporates an additional geometric loss term based on an eigenvalue-derived 3D shape feature into the optimization process of 3DGS.
  arXiv  Detail & Related papers  (2025-01-29T13:40:25Z)
• AGS-Mesh: Adaptive Gaussian Splatting and Meshing with Geometric Priors for Indoor Room Reconstruction Using Smartphones [19.429461194706786]
  We propose an approach for joint surface depth and normal refinement of Gaussian Splatting methods for accurate 3D reconstruction of indoor scenes.<n>Our filtering strategy and optimization design demonstrate significant improvements in both mesh estimation and novel-view synthesis.
  arXiv  Detail & Related papers  (2024-11-28T17:04:32Z)
• Structure Consistent Gaussian Splatting with Matching Prior for Few-shot Novel View Synthesis [28.3325478008559]
  We propose SCGaussian, a Structure Consistent Gaussian Splatting method using matching priors to learn 3D consistent scene structure. We optimize the scene structure in two folds: rendering geometry and, more importantly, the position of Gaussian primitives. Experiments on forward-facing, surrounding, and complex large scenes show the effectiveness of our approach with state-of-the-art performance and high efficiency.
  arXiv  Detail & Related papers  (2024-11-06T03:28:06Z)
• Trim 3D Gaussian Splatting for Accurate Geometry Representation [72.00970038074493]
  We introduce Trim 3D Gaussian Splatting (TrimGS) to reconstruct accurate 3D geometry from images. Our experimental and theoretical analyses reveal that a relatively small Gaussian scale is a non-negligible factor in representing and optimizing the intricate details. When combined with the original 3DGS and the state-of-the-art 2DGS, TrimGS consistently yields more accurate geometry and higher perceptual quality.
  arXiv  Detail & Related papers  (2024-06-11T17:34:46Z)
• GVGEN: Text-to-3D Generation with Volumetric Representation [89.55687129165256]
  3D Gaussian splatting has emerged as a powerful technique for 3D reconstruction and generation, known for its fast and high-quality rendering capabilities. This paper introduces a novel diffusion-based framework, GVGEN, designed to efficiently generate 3D Gaussian representations from text input.
  arXiv  Detail & Related papers  (2024-03-19T17:57:52Z)
• Leveraging Non-uniformity in First-order Non-convex Optimization [93.6817946818977]
  Non-uniform refinement of objective functions leads to emphNon-uniform Smoothness (NS) and emphNon-uniform Lojasiewicz inequality (NL) New definitions inspire new geometry-aware first-order methods that converge to global optimality faster than the classical $Omega (1/t2)$ lower bounds.
  arXiv  Detail & Related papers  (2021-05-13T04:23:07Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.