CityGaussian: Real-time High-quality Large-Scale Scene Rendering with Gaussians
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910530003795968 |
|---|---|
| author | Liu, Yang Guan, He Luo, Chuanchen Fan, Lue Wang, Naiyan Peng, Junran Zhang, Zhaoxiang |
| author_facet | Liu, Yang Guan, He Luo, Chuanchen Fan, Lue Wang, Naiyan Peng, Junran Zhang, Zhaoxiang |
| contents | The advancement of real-time 3D scene reconstruction and novel view synthesis has been significantly propelled by 3D Gaussian Splatting (3DGS). However, effectively training large-scale 3DGS and rendering it in real-time across various scales remains challenging. This paper introduces CityGaussian (CityGS), which employs a novel divide-and-conquer training approach and Level-of-Detail (LoD) strategy for efficient large-scale 3DGS training and rendering. Specifically, the global scene prior and adaptive training data selection enables efficient training and seamless fusion. Based on fused Gaussian primitives, we generate different detail levels through compression, and realize fast rendering across various scales through the proposed block-wise detail levels selection and aggregation strategy. Extensive experimental results on large-scale scenes demonstrate that our approach attains state-of-theart rendering quality, enabling consistent real-time rendering of largescale scenes across vastly different scales. Our project page is available at https://dekuliutesla.github.io/citygs/. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_01133 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | CityGaussian: Real-time High-quality Large-Scale Scene Rendering with Gaussians Liu, Yang Guan, He Luo, Chuanchen Fan, Lue Wang, Naiyan Peng, Junran Zhang, Zhaoxiang Computer Vision and Pattern Recognition The advancement of real-time 3D scene reconstruction and novel view synthesis has been significantly propelled by 3D Gaussian Splatting (3DGS). However, effectively training large-scale 3DGS and rendering it in real-time across various scales remains challenging. This paper introduces CityGaussian (CityGS), which employs a novel divide-and-conquer training approach and Level-of-Detail (LoD) strategy for efficient large-scale 3DGS training and rendering. Specifically, the global scene prior and adaptive training data selection enables efficient training and seamless fusion. Based on fused Gaussian primitives, we generate different detail levels through compression, and realize fast rendering across various scales through the proposed block-wise detail levels selection and aggregation strategy. Extensive experimental results on large-scale scenes demonstrate that our approach attains state-of-theart rendering quality, enabling consistent real-time rendering of largescale scenes across vastly different scales. Our project page is available at https://dekuliutesla.github.io/citygs/. |
| title | CityGaussian: Real-time High-quality Large-Scale Scene Rendering with Gaussians |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2404.01133 |