UrbanGS: A Scalable and Efficient Architecture for Geometrically Accurate Large-Scene Reconstruction

2026-02-02

AI Summary

This work addresses the challenges of scaling 3D Gaussian Splatting to large-scale urban scenes, where geometric inconsistency, excessive memory consumption, and poor computational scalability hinder performance. To overcome these limitations, the authors propose UrbanGS, a novel framework that integrates external depth supervision through depth-consistent D-Normal regularization and an adaptive confidence-weighting mechanism to enhance geometric accuracy. Furthermore, UrbanGS introduces Spatially Adaptive Gaussian Pruning (SAGP) and a unified partitioned view assignment strategy to effectively suppress boundary artifacts and significantly reduce memory footprint. Extensive experiments demonstrate that UrbanGS consistently outperforms existing methods across multiple urban scene datasets, achieving systematic improvements in rendering quality, geometric fidelity, and scalability.

Technology Category

• Computer Vision: 3D Computer Vision
• Search and Optimization: Sampling/Simulation-based Search
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

While 3D Gaussian Splatting (3DGS) enables high-quality, real-time rendering for bounded scenes, its extension to large-scale urban environments gives rise to critical challenges in terms of geometric consistency, memory efficiency, and computational scalability. To address these issues, we present UrbanGS, a scalable reconstruction framework that effectively tackles these challenges for city-scale applications. First, we propose a Depth-Consistent D-Normal Regularization module. Unlike existing approaches that rely solely on monocular normal estimators, which can effectively update rotation parameters yet struggle to update position parameters, our method integrates D-Normal constraints with external depth supervision. This allows for comprehensive updates of all geometric parameters. By further incorporating an adaptive confidence weighting mechanism based on gradient consistency and inverse depth deviation, our approach significantly enhances multi-view depth alignment and geometric coherence, which effectively resolves the issue of geometric accuracy in complex large-scale scenes. To improve scalability, we introduce a Spatially Adaptive Gaussian Pruning (SAGP) strategy, which dynamically adjusts Gaussian density based on local geometric complexity and visibility to reduce redundancy. Additionally, a unified partitioning and view assignment scheme is designed to eliminate boundary artifacts and optimize computational load. Extensive experiments on multiple urban datasets demonstrate that UrbanGS achieves superior performance in rendering quality, geometric accuracy, and memory efficiency, providing a systematic solution for high-fidelity large-scale scene reconstruction.

Problem

• 3D Gaussian Splatting
• large-scale reconstruction
• geometric consistency
• memory efficiency
• computational scalability

Innovation

• 3D Gaussian Splatting
• Depth-Consistent D-Normal Regularization
• Spatially Adaptive Gaussian Pruning
• Large-Scale Scene Reconstruction
• Geometric Consistency