Horizon-GS: Unified 3D Gaussian Splatting for Large-Scale Aerial-to-Ground Scenes

TL;DR

Horizon-GS is a unified 3D Gaussian Splatting method that enables seamless reconstruction and rendering of large-scale aerial and ground scenes, facilitating immersive view exploration in urban environments.

Contribution

We propose Horizon-GS, a novel Gaussian Splatting-based approach for unified aerial and ground scene reconstruction, along with a new training strategy and a comprehensive dataset.

Findings

Effective in diverse urban scenes

High-fidelity scene generation

Overcomes viewpoint discrepancies

Abstract

Seamless integration of both aerial and street view images remains a significant challenge in neural scene reconstruction and rendering. Existing methods predominantly focus on single domain, limiting their applications in immersive environments, which demand extensive free view exploration with large view changes both horizontally and vertically. We introduce Horizon-GS, a novel approach built upon Gaussian Splatting techniques, tackles the unified reconstruction and rendering for aerial and street views. Our method addresses the key challenges of combining these perspectives with a new training strategy, overcoming viewpoint discrepancies to generate high-fidelity scenes. We also curate a high-quality aerial-to-ground views dataset encompassing both synthetic and real-world scene to advance further research. Experiments across diverse urban scene datasets confirm the effectiveness of our method.