BOGausS: Better Optimized Gaussian Splatting

Stéphane Pateux 1 , Matthieu Gendrin 1,2 , Luce Morin 2 , Théo Ladune 1 , Xiaoran Jiang 2 · 2025

Read BOGausS: Better Optimized Gaussian Splatting by Stéphane Pateux 1 , Matthieu Gendrin 1,2 , Luce Morin 2 , Théo Ladune 1 , Xiaoran Jiang 2 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-01844.txt.

Description

Stéphane Pateux and co-authors make 3D Gaussian Splatting models far smaller without losing picture quality, in work from 2025. The technique renders new views of a scene quickly and sharply, but its models are still bulky next to some rivals.

After a careful study of how the models are trained, the team proposes BOGausS, a retuned optimization recipe. The resulting models can be up to ten times lighter than the original method's, with no visible quality loss. That improvement, the authors argue, pushes Gaussian Splatting ahead of the state of the art for practical novel-view rendering, where file size matters for storage and streaming.

Catalog record

CREATORStéphane Pateux 1 , Matthieu Gendrin 1,2 , Luce Morin 2 , Théo Ladune 1 , Xiaoran Jiang 2
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE23 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-01844

Browse the full library · About the Tower of Babel