Improving Densification in 3D Gaussian Splatting for High-Fidelity Rendering

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Deng, Xiaobin, Diao, Changyu, Li, Min, Yu, Ruohan, Xu, Duanqing
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866909739614470144
author Deng, Xiaobin
Diao, Changyu
Li, Min
Yu, Ruohan
Xu, Duanqing
author_facet Deng, Xiaobin
Diao, Changyu
Li, Min
Yu, Ruohan
Xu, Duanqing
contents Although 3D Gaussian Splatting (3DGS) has achieved impressive performance in real-time rendering, its densification strategy often results in suboptimal reconstruction quality. In this work, we present a comprehensive improvement to the densification pipeline of 3DGS from three perspectives: when to densify, how to densify, and how to mitigate overfitting. Specifically, we propose an Edge-Aware Score to effectively select candidate Gaussians for splitting. We further introduce a Long-Axis Split strategy that reduces geometric distortions introduced by clone and split operations. To address overfitting, we design a set of techniques, including Recovery-Aware Pruning, Multi-step Update, and Growth Control. Our method enhances rendering fidelity without introducing additional training or inference overhead, achieving state-of-the-art performance with fewer Gaussians.
format Preprint
id arxiv_https___arxiv_org_abs_2508_12313
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Improving Densification in 3D Gaussian Splatting for High-Fidelity Rendering
Deng, Xiaobin
Diao, Changyu
Li, Min
Yu, Ruohan
Xu, Duanqing
Computer Vision and Pattern Recognition
Although 3D Gaussian Splatting (3DGS) has achieved impressive performance in real-time rendering, its densification strategy often results in suboptimal reconstruction quality. In this work, we present a comprehensive improvement to the densification pipeline of 3DGS from three perspectives: when to densify, how to densify, and how to mitigate overfitting. Specifically, we propose an Edge-Aware Score to effectively select candidate Gaussians for splitting. We further introduce a Long-Axis Split strategy that reduces geometric distortions introduced by clone and split operations. To address overfitting, we design a set of techniques, including Recovery-Aware Pruning, Multi-step Update, and Growth Control. Our method enhances rendering fidelity without introducing additional training or inference overhead, achieving state-of-the-art performance with fewer Gaussians.
title Improving Densification in 3D Gaussian Splatting for High-Fidelity Rendering
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2508.12313