SkySplat: Generalizable 3D Gaussian Splatting from Multi-Temporal Sparse Satellite Images

TL;DR

SkySplat introduces a self-supervised 3D reconstruction framework for multi-temporal satellite images, integrating RPC models and novel modules to improve accuracy, efficiency, and generalization without ground-truth height maps.

Contribution

It presents SkySplat, a new method that combines RPC models with 3D Gaussian Splatting, featuring modules for transient object masking and multi-view consistency, enhancing satellite image reconstruction.

Findings

Achieves 86x speedup over EOGS with higher accuracy.

Reduces MAE from 13.18 m to 1.80 m on DFC19 dataset.

Demonstrates strong cross-dataset generalization.

Abstract

Three-dimensional scene reconstruction from sparse-view satellite images is a long-standing and challenging task. While 3D Gaussian Splatting (3DGS) and its variants have recently attracted attention for its high efficiency, existing methods remain unsuitable for satellite images due to incompatibility with rational polynomial coefficient (RPC) models and limited generalization capability. Recent advances in generalizable 3DGS approaches show potential, but they perform poorly on multi-temporal sparse satellite images due to limited geometric constraints, transient objects, and radiometric inconsistencies. To address these limitations, we propose SkySplat, a novel self-supervised framework that integrates the RPC model into the generalizable 3DGS pipeline, enabling more effective use of sparse geometric cues for improved reconstruction. SkySplat relies only on RGB images and radiometric-robust relative height supervision, thereby eliminating the need for ground-truth height maps. Key components include a Cross-Self Consistency Module (CSCM), which mitigates transient object interference via consistency-based masking, and a multi-view consistency aggregation strategy that refines reconstruction results. Compared to per-scene optimization methods, SkySplat achieves an 86 times speedup over EOGS with higher accuracy. It also outperforms generalizable 3DGS baselines, reducing MAE from 13.18 m to 1.80 m on the DFC19 dataset significantly, and demonstrates strong cross-dataset generalization on the MVS3D benchmark. The is available at https://github.com/NanCheng2001/SkySplat-main