Skip to content

Gradient-Driven Point Allocation for Depth-Guided 3D Gaussian Splatting in Indoor Scenes

Oct 2026 · SAE technical paper series · 0 citations · 7 references

TL;DR

A lightweight gradient-driven point allocation strategy for depth-guided Gaussian splatting that bypasses traditional Structure-from-Motion initialization and incorporates a bidirectional re-projection check to eliminate transient noise and occlusions, ensuring geometric consistency is proposed.

Abstract

High-precision 3D environmental perception is crucial for autonomous systems to navigate and interact effectively within complex indoor environments. However, achieving high-granularity 3D reconstruction and high-fidelity rendering remains challenging. Although 3D Gaussian Splatting (3DGS) enables efficient photorealistic rendering, its geometric inconsistency limits its ability to reconstruct high-accuracy scenes. Standard 3DGS relies on sparse Structure-from-Motion (SfM) initialization, which often leads to uneven point distributions and floaters, particularly in weak-texture areas like plain walls. Furthermore, existing methods based on depth regularization or geometric constraints usually fail to address the fundamental sparsity of the initialization. To address these issues, we propose a lightweight gradient-driven point allocation strategy for depth-guided Gaussian splatting that bypasses traditional SfM. By utilizing RGB-D priors, we obtain a high-quality, structurally continuous initial Gaussian distribution through multi-view depth back-projection and texture-aware density control. Our method incorporates a bidirectional re-projection check to eliminate transient noise and occlusions, ensuring geometric consistency. We introduce a dual-grid allocation scheme that adaptively distributes Gaussians based on spatial frequency, mandating density in high-gradient regions while enforcing sparsity in flat areas. Quantitative and qualitative experiments on the ScanNet++ dataset demonstrate that our method achieves superior reconstruction accuracy and rendering quality. Notably, our approach achieves state-of-the-art performance with a significantly reduced Gaussian count, decreasing from 1.70M to 0.39M in representative scenes, thereby mitigating unnecessary point cloud bloat and reducing memory overhead without sacrificing structural integrity.

View source

Similar papers

Book Open access Nov 2026

MPP-GS: Multi-Granularity Planar Prior Guided Gaussian Splatting for Indoor Scene Reconstruction

3D Gaussian Splatting (3DGS) has recently achieved impressive performance in novel view synthesis and real-time rendering. However, its geometric reconstruction quality in indoor scenes remains limited, especially in large weak-texture regions such as walls, ceilings, and floors, where sparse initialization and purely...

Liang-Ting Wang, Hong-Xing Qin · 0 citations
Book Open access Nov 2026

PD-GS: Prior-based Decoupled Optimization for Unbounded Scene Gaussian Splatting

Prized for its high-quality real-time rendering, 3D Gaussian Splatting is a key novel view synthesis technology for fields such as virtual reality and autonomous driving, where it is widely used for outdoor scene reconstruction. However, extensive unbounded and texture-scarce regions in outdoor scenes, such as the sky,...

Jian Xiong, Qing Zhang, Jiu-Cheng Xie et al. · 0 citations
Sep 2026

D3GS: Depth, DINO, and RGB Diffusion Co-Guided 3D Gaussian Splatting for Sparse-View Reconstruction

D3GS is proposed, a Depth-DINO-Diffusion guided sparse-view Gaussian reconstruction framework that jointly enhances geometry and appearance and injects generative priors into an iterative optimization strategy, enhancing high-frequency geometric and appearance details within the Gaussian representation.

Yun-Qi Gao, Zhan-Feng Liao, Han-Zhang Tu et al. · 0 citations
Preprint Nov 2026

OctreeSplat: Hierarchical Octree-Guided Feed-Forward Gaussian Splatting

3D Gaussian Splatting (3DGS) has shown considerable potential for novel view synthesis (NVS) and 3D reconstruction. Feed-forward methods avoid costly per-scene optimization, but fixed image or spatial resolutions limit their ability to allocate primitives according to local complexity. We present OctreeSplat, a hierarc...

Yi-Bin Zhao, Yi-Han Pan, Wen-Li Yang et al. · 0 citations
Oct 2026

G 2 -SLAM: Geometrically-guided 3D Gaussian Splatting SLAM via robust loop validation and anisotropic shaping

Monocular 3D Gaussian Splatting (3DGS) SLAM estimates camera motion and dense scene appearance from a single RGB stream, but sparse views make both estimates sensitive to depth uncertainty and false loop closures. We present G²-SLAM, a geometry-guided monocular 3DGS SLAM system that combines depth-Hessian-based loop...

Xiong Gu, Ming-Li Sun, You-Hui Zhang et al. · 0 citations
Preprint Oct 2026

3D Gaussian Splatting for Fine-grained Urban Scene Reconstruction via Semantic-guided Density Control and Loss Design

3D Gaussian Splatting (3DGS) has been becoming a promising 3D reconstruction technique because of its advantage of high-fidelity rendering effect with high efficiency. However, it still suffers from coarse density control and semantically blind loss functions when reconstructing the fine-grained structures in complex u...

Yue-Ling Liao, San Jiang, Xing Zhang et al. · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.