VastGaussian: Vast 3D Gaussians for Large Scene Reconstruction
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866914693039259648 |
|---|---|
| author | Lin, Jiaqi Li, Zhihao Tang, Xiao Liu, Jianzhuang Liu, Shiyong Liu, Jiayue Lu, Yangdi Wu, Xiaofei Xu, Songcen Yan, Youliang Yang, Wenming |
| author_facet | Lin, Jiaqi Li, Zhihao Tang, Xiao Liu, Jianzhuang Liu, Shiyong Liu, Jiayue Lu, Yangdi Wu, Xiaofei Xu, Songcen Yan, Youliang Yang, Wenming |
| contents | Existing NeRF-based methods for large scene reconstruction often have limitations in visual quality and rendering speed. While the recent 3D Gaussian Splatting works well on small-scale and object-centric scenes, scaling it up to large scenes poses challenges due to limited video memory, long optimization time, and noticeable appearance variations. To address these challenges, we present VastGaussian, the first method for high-quality reconstruction and real-time rendering on large scenes based on 3D Gaussian Splatting. We propose a progressive partitioning strategy to divide a large scene into multiple cells, where the training cameras and point cloud are properly distributed with an airspace-aware visibility criterion. These cells are merged into a complete scene after parallel optimization. We also introduce decoupled appearance modeling into the optimization process to reduce appearance variations in the rendered images. Our approach outperforms existing NeRF-based methods and achieves state-of-the-art results on multiple large scene datasets, enabling fast optimization and high-fidelity real-time rendering. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_17427 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | VastGaussian: Vast 3D Gaussians for Large Scene Reconstruction Lin, Jiaqi Li, Zhihao Tang, Xiao Liu, Jianzhuang Liu, Shiyong Liu, Jiayue Lu, Yangdi Wu, Xiaofei Xu, Songcen Yan, Youliang Yang, Wenming Computer Vision and Pattern Recognition Existing NeRF-based methods for large scene reconstruction often have limitations in visual quality and rendering speed. While the recent 3D Gaussian Splatting works well on small-scale and object-centric scenes, scaling it up to large scenes poses challenges due to limited video memory, long optimization time, and noticeable appearance variations. To address these challenges, we present VastGaussian, the first method for high-quality reconstruction and real-time rendering on large scenes based on 3D Gaussian Splatting. We propose a progressive partitioning strategy to divide a large scene into multiple cells, where the training cameras and point cloud are properly distributed with an airspace-aware visibility criterion. These cells are merged into a complete scene after parallel optimization. We also introduce decoupled appearance modeling into the optimization process to reduce appearance variations in the rendered images. Our approach outperforms existing NeRF-based methods and achieves state-of-the-art results on multiple large scene datasets, enabling fast optimization and high-fidelity real-time rendering. |
| title | VastGaussian: Vast 3D Gaussians for Large Scene Reconstruction |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2402.17427 |