DGTR: Distributed Gaussian Turbo-Reconstruction for Sparse-View Vast Scenes

TL;DR

DGTR introduces a distributed Gaussian-based framework for efficient, high-quality sparse-view reconstruction of large scenes, significantly reducing training time and improving scalability in novel-view synthesis tasks.

Contribution

The paper proposes a novel distributed Gaussian reconstruction method that enables fast, scalable, and high-quality scene reconstruction from sparse views, addressing limitations of existing methods.

Findings

Achieves high-quality large-scale scene reconstruction within minutes.

Outperforms existing approaches in speed and scalability.

Effectively integrates synthetic views and depth priors for enhanced results.

Abstract

Novel-view synthesis (NVS) 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. Synthetic views and depth priors are 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 [https://3d-aigc.github.io/DGTR].