Saved in:
Bibliographic Details
Main Authors: Windisch, Felix, Köhler, Thomas, Radl, Lukas, D'Urso, Mattia, Steiner, Michael, Schmalstieg, Dieter, Steinberger, Markus
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2507.01110
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911631320023040
author Windisch, Felix
Köhler, Thomas
Radl, Lukas
D'Urso, Mattia
Steiner, Michael
Schmalstieg, Dieter
Steinberger, Markus
author_facet Windisch, Felix
Köhler, Thomas
Radl, Lukas
D'Urso, Mattia
Steiner, Michael
Schmalstieg, Dieter
Steinberger, Markus
contents Gaussian Splatting has emerged as a high-performance technique for novel view synthesis, enabling real-time rendering and high-quality reconstruction of small scenes. However, scaling to larger environments has so far relied on partitioning the scene into chunks -- a strategy that introduces artifacts at chunk boundaries, complicates training across varying scales, and is poorly suited to unstructured scenarios such as city-scale flyovers combined with street-level views. Moreover, rendering remains fundamentally limited by GPU memory, as all visible chunks must reside in VRAM simultaneously. We introduce A LoD of Gaussians, a framework for training and rendering ultra-large-scale Gaussian scenes on a single consumer-grade GPU -- without partitioning. Our method stores the full scene out-of-core (e.g., in CPU memory) and trains a Level-of-Detail (LoD) representation directly, dynamically streaming only the relevant Gaussians. A hybrid data structure combining Gaussian hierarchies with Sequential Point Trees enables efficient, view-dependent LoD selection, while a lightweight caching and view scheduling system exploits temporal coherence to support real-time streaming and rendering. Together, these innovations enable seamless multi-scale reconstruction and interactive visualization of complex scenes -- from broad aerial views to fine-grained ground-level details.
format Preprint
id arxiv_https___arxiv_org_abs_2507_01110
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A LoD of Gaussians: Unified Training and Rendering for Ultra-Large Scale Reconstruction with External Memory
Windisch, Felix
Köhler, Thomas
Radl, Lukas
D'Urso, Mattia
Steiner, Michael
Schmalstieg, Dieter
Steinberger, Markus
Graphics
Machine Learning
Gaussian Splatting has emerged as a high-performance technique for novel view synthesis, enabling real-time rendering and high-quality reconstruction of small scenes. However, scaling to larger environments has so far relied on partitioning the scene into chunks -- a strategy that introduces artifacts at chunk boundaries, complicates training across varying scales, and is poorly suited to unstructured scenarios such as city-scale flyovers combined with street-level views. Moreover, rendering remains fundamentally limited by GPU memory, as all visible chunks must reside in VRAM simultaneously. We introduce A LoD of Gaussians, a framework for training and rendering ultra-large-scale Gaussian scenes on a single consumer-grade GPU -- without partitioning. Our method stores the full scene out-of-core (e.g., in CPU memory) and trains a Level-of-Detail (LoD) representation directly, dynamically streaming only the relevant Gaussians. A hybrid data structure combining Gaussian hierarchies with Sequential Point Trees enables efficient, view-dependent LoD selection, while a lightweight caching and view scheduling system exploits temporal coherence to support real-time streaming and rendering. Together, these innovations enable seamless multi-scale reconstruction and interactive visualization of complex scenes -- from broad aerial views to fine-grained ground-level details.
title A LoD of Gaussians: Unified Training and Rendering for Ultra-Large Scale Reconstruction with External Memory
topic Graphics
Machine Learning
url https://arxiv.org/abs/2507.01110