HUG: Hierarchical Urban Gaussian Splatting with Block-Based Reconstruction

2025-04-23

AI Summary

To address data redundancy, boundary artifacts, and excessive computational overhead in Gaussian splatting for large-scale urban 3D scene reconstruction and rendering, this paper proposes a hierarchical neural Gaussian representation coupled with a block-adaptive reconstruction architecture. Our method introduces, for the first time, inter-block decoupled training and a boundary-redundancy suppression mechanism, integrating hierarchical neural radiance field modeling, sparse voxel-guided optimization, and a boundary-aware loss function. This design ensures global geometric consistency while significantly enhancing local detail fidelity. Evaluated on mainstream urban benchmarks, our approach achieves state-of-the-art performance: +1.8 dB PSNR and +0.022 SSIM improvements over prior methods; 37% reduction in training memory consumption; and 2.1× speedup in inference time.

Technology Category

• Computer Vision: 3D Computer Vision
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

As urban 3D scenes become increasingly complex and the demand for high-quality rendering grows, efficient scene reconstruction and rendering techniques become crucial. We present HUG, a novel approach to address inefficiencies in handling large-scale urban environments and intricate details based on 3D Gaussian splatting. Our method optimizes data partitioning and the reconstruction pipeline by incorporating a hierarchical neural Gaussian representation. We employ an enhanced block-based reconstruction pipeline focusing on improving reconstruction quality within each block and reducing the need for redundant training regions around block boundaries. By integrating neural Gaussian representation with a hierarchical architecture, we achieve high-quality scene rendering at a low computational cost. This is demonstrated by our state-of-the-art results on public benchmarks, which prove the effectiveness and advantages in large-scale urban scene representation.

Problem

• Efficient reconstruction of complex urban 3D scenes
• Handling large-scale urban environments with intricate details
• Reducing computational cost while maintaining rendering quality

Innovation

• Hierarchical neural Gaussian representation for urban scenes
• Block-based reconstruction pipeline for efficiency
• Enhanced quality with reduced redundant training regions