DGTR: Distributed Gaussian Turbo-Reconstruction for Sparse-View Vast Scenes
Authors:
- 19 May 2025
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 207-213
Abstract
Novel-view synthesis approaches play a critical role in vast scene reconstruction. However, these methods rely heavily on dense image inputs and prolonged training times, making them unsuitable where computational resources are limited. Additionally, few-shot methods often struggle with poor reconstruction quality in vast environments. This paper presents DGTR, a novel distributed framework for efficient Gaussian reconstruction for sparse-view vast scenes. Our approach divides the scene into regions, processed independently by drones with sparse image inputs. Using a feed-forward Gaussian model, we predict high-quality Gaussian primitives, followed by a global alignment algorithm to ensure geometric consistency. Depth priors is incorporated to further enhance training, while a distillation-based model aggregation mechanism enables efficient reconstruction. Our method achieves high-quality large-scale scene reconstruction and novel-view synthesis in significantly reduced training times, outperforming existing approaches in both speed and scalability. We demonstrate the effectiveness of our framework on vast aerial scenes, achieving high-quality results within minutes. Code will released on our project page https://3d-aigc.github.io/DGTR.Keywords
This publication has 35 references indexed in Scilit:
- Flash3D: Feed-Forward Generalisable 3D Scene Reconstruction from a Single Image2025
- HO-Gaussian: Hybrid Optimization of 3D Gaussian Splatting for Urban Scenes2024
- CityGaussian: Real-Time High-Quality Large-Scale Scene Rendering with Gaussians2024
- MVSplat: Efficient 3D Gaussian Splatting from Sparse Multi-view Images2024
- FSGS: Real-Time Few-Shot View Synthesis Using Gaussian Splatting2024
- Efficient large-scale scene representation with a hybrid of high-resolution grid and plane featuresPattern Recognition, 2024
- Drone-NeRF: Efficient NeRF based 3D scene reconstruction for large-scale drone surveyImage and Vision Computing, 2024
- Capturing, Reconstructing, and Simulating: The UrbanScene3D Dataset2022
- BungeeNeRF: Progressive Neural Radiance Field for Extreme Multi-scale Scene Rendering2022
- An Automated Method for Large-Scale, Ground-Based City Model AcquisitionInternational Journal of Computer Vision, 2004