Skip to content

Official Implementation of Arxiv 2025 paper 'FreeSplat++: Generalizable 3D Gaussian Splatting for Efficient Indoor Scene Reconstruction'

Notifications You must be signed in to change notification settings

wangys16/FreeSplatPP

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 

History

32 Commits
Β 
Β 
Β 
Β 

Repository files navigation

FreeSplat++: Generalizable 3D Gaussian Splatting for Efficient Indoor Scene Reconstruction

Yunsong Wang, Tianxin Huang, Hanlin Chen, Gim Hee Lee

Arxiv 2025

arXiv Project page

demo.mp4

Logo

Logo

πŸ“ƒ Abstract

FreeSplat++ is an enhanced version of FreeSplat to solve the feed-forward whole scene reconstruction challenge, in which we carefully design a fusion module to mitigate 3D Gaussian redundancy, and a Weighted Floater Removal scheme to significantly reduce floater Gaussians in whole scene reconstruction. Furthermore, we also investigate a depth-regularized fine-tuning step to further enhance the rendering quality while maintaining the geometric accuracy. Empirical results demonstrate our significant improvements over existing feed-forward methods and per-scene optimization method in terms of rendering quality, geometric accuracy and efficiency.

🌏 News

The code will be released recently.

πŸ–Š BibTeX

If you find our work helpful, please consider citing our papers. Thank you!

@article{wang2025freesplat++,
  title={FreeSplat++: Generalizable 3D Gaussian Splatting for Efficient Indoor Scene Reconstruction},
  author={Wang, Yunsong and Huang, Tianxin and Chen, Hanlin and Lee, Gim Hee},
  journal={arXiv preprint arXiv:2503.22986},
  year={2025}
}
@article{wang2024freesplat,
  title={FreeSplat: Generalizable 3D Gaussian Splatting Towards Free-View Synthesis of Indoor Scenes},
  author={Wang, Yunsong and Huang, Tianxin and Chen, Hanlin and Lee, Gim Hee},
  journal={arXiv preprint arXiv:2405.17958},
  year={2024}
}

About

Official Implementation of Arxiv 2025 paper 'FreeSplat++: Generalizable 3D Gaussian Splatting for Efficient Indoor Scene Reconstruction'

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published