DashGaussian: Optimizing 3D Gaussian Splatting in 200 Seconds

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Youyu, Jiang, Junjun, Jiang, Kui, Tang, Xiao, Li, Zhihao, Liu, Xianming, Nie, Yinyu
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912294503448576
author Chen, Youyu
Jiang, Junjun
Jiang, Kui
Tang, Xiao
Li, Zhihao
Liu, Xianming
Nie, Yinyu
author_facet Chen, Youyu
Jiang, Junjun
Jiang, Kui
Tang, Xiao
Li, Zhihao
Liu, Xianming
Nie, Yinyu
contents 3D Gaussian Splatting (3DGS) renders pixels by rasterizing Gaussian primitives, where the rendering resolution and the primitive number, concluded as the optimization complexity, dominate the time cost in primitive optimization. In this paper, we propose DashGaussian, a scheduling scheme over the optimization complexity of 3DGS that strips redundant complexity to accelerate 3DGS optimization. Specifically, we formulate 3DGS optimization as progressively fitting 3DGS to higher levels of frequency components in the training views, and propose a dynamic rendering resolution scheme that largely reduces the optimization complexity based on this formulation. Besides, we argue that a specific rendering resolution should cooperate with a proper primitive number for a better balance between computing redundancy and fitting quality, where we schedule the growth of the primitives to synchronize with the rendering resolution. Extensive experiments show that our method accelerates the optimization of various 3DGS backbones by 45.7% on average while preserving the rendering quality.
format Preprint
id arxiv_https___arxiv_org_abs_2503_18402
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle DashGaussian: Optimizing 3D Gaussian Splatting in 200 Seconds
Chen, Youyu
Jiang, Junjun
Jiang, Kui
Tang, Xiao
Li, Zhihao
Liu, Xianming
Nie, Yinyu
Computer Vision and Pattern Recognition
3D Gaussian Splatting (3DGS) renders pixels by rasterizing Gaussian primitives, where the rendering resolution and the primitive number, concluded as the optimization complexity, dominate the time cost in primitive optimization. In this paper, we propose DashGaussian, a scheduling scheme over the optimization complexity of 3DGS that strips redundant complexity to accelerate 3DGS optimization. Specifically, we formulate 3DGS optimization as progressively fitting 3DGS to higher levels of frequency components in the training views, and propose a dynamic rendering resolution scheme that largely reduces the optimization complexity based on this formulation. Besides, we argue that a specific rendering resolution should cooperate with a proper primitive number for a better balance between computing redundancy and fitting quality, where we schedule the growth of the primitives to synchronize with the rendering resolution. Extensive experiments show that our method accelerates the optimization of various 3DGS backbones by 45.7% on average while preserving the rendering quality.
title DashGaussian: Optimizing 3D Gaussian Splatting in 200 Seconds
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2503.18402