GS4Buildings is a prior-guided Gaussian Splatting framework for robust and complete 3D building reconstruction, leveraging LoD2 semantic building models to improve completeness and geometric accuracy in urban scenes.
- SfM-free initialization using LoD2 building models and known camera parameters.
- Depth and normal priors generated from semantic models for geometry-aware optimization.
- Building-focused mode for efficient and compact reconstructions.
- Applicable to smart cities and digital twin applications.
📢 Code coming soon!
This repository will contain the implementation, example datasets, and usage instructions described in our paper:
Q. Zhang, O. Wysocki, B. Jutzi. GS4Buildings: Prior-Guided Gaussian Splatting for 3D Building Reconstruction, ISPRS 3D GeoInfo & Smart Data, Smart Cities 2025.
If you find this work useful, please consider citing:
@inproceedings{zhang2025gs4buildings,
title = {GS4Buildings: Prior-Guided Gaussian Splatting for 3D Building Reconstruction},
author = {Zhang, Qilin and Wysocki, Olaf and Jutzi, Boris},
booktitle = {ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences},
year = {2025}
}