CRUISE: Cooperative Reconstruction and Editing in V2X Scenarios using Gaussian Splatting
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866918103591419904 |
|---|---|
| author | Xu, Haoran Zhang, Saining Li, Peishuo Ye, Baijun Chen, Xiaoxue Gao, Huan-ang Zheng, Jv Song, Xiaowei Peng, Ziqiao Miao, Run Jia, Jinrang Shi, Yifeng Yi, Guangqi Zhao, Hang Tang, Hao Li, Hongyang Yu, Kaicheng Zhao, Hao |
| author_facet | Xu, Haoran Zhang, Saining Li, Peishuo Ye, Baijun Chen, Xiaoxue Gao, Huan-ang Zheng, Jv Song, Xiaowei Peng, Ziqiao Miao, Run Jia, Jinrang Shi, Yifeng Yi, Guangqi Zhao, Hang Tang, Hao Li, Hongyang Yu, Kaicheng Zhao, Hao |
| contents | Vehicle-to-everything (V2X) communication plays a crucial role in autonomous driving, enabling cooperation between vehicles and infrastructure. While simulation has significantly contributed to various autonomous driving tasks, its potential for data generation and augmentation in V2X scenarios remains underexplored. In this paper, we introduce CRUISE, a comprehensive reconstruction-and-synthesis framework designed for V2X driving environments. CRUISE employs decomposed Gaussian Splatting to accurately reconstruct real-world scenes while supporting flexible editing. By decomposing dynamic traffic participants into editable Gaussian representations, CRUISE allows for seamless modification and augmentation of driving scenes. Furthermore, the framework renders images from both ego-vehicle and infrastructure views, enabling large-scale V2X dataset augmentation for training and evaluation. Our experimental results demonstrate that: 1) CRUISE reconstructs real-world V2X driving scenes with high fidelity; 2) using CRUISE improves 3D detection across ego-vehicle, infrastructure, and cooperative views, as well as cooperative 3D tracking on the V2X-Seq benchmark; and 3) CRUISE effectively generates challenging corner cases. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_18473 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | CRUISE: Cooperative Reconstruction and Editing in V2X Scenarios using Gaussian Splatting Xu, Haoran Zhang, Saining Li, Peishuo Ye, Baijun Chen, Xiaoxue Gao, Huan-ang Zheng, Jv Song, Xiaowei Peng, Ziqiao Miao, Run Jia, Jinrang Shi, Yifeng Yi, Guangqi Zhao, Hang Tang, Hao Li, Hongyang Yu, Kaicheng Zhao, Hao Computer Vision and Pattern Recognition Vehicle-to-everything (V2X) communication plays a crucial role in autonomous driving, enabling cooperation between vehicles and infrastructure. While simulation has significantly contributed to various autonomous driving tasks, its potential for data generation and augmentation in V2X scenarios remains underexplored. In this paper, we introduce CRUISE, a comprehensive reconstruction-and-synthesis framework designed for V2X driving environments. CRUISE employs decomposed Gaussian Splatting to accurately reconstruct real-world scenes while supporting flexible editing. By decomposing dynamic traffic participants into editable Gaussian representations, CRUISE allows for seamless modification and augmentation of driving scenes. Furthermore, the framework renders images from both ego-vehicle and infrastructure views, enabling large-scale V2X dataset augmentation for training and evaluation. Our experimental results demonstrate that: 1) CRUISE reconstructs real-world V2X driving scenes with high fidelity; 2) using CRUISE improves 3D detection across ego-vehicle, infrastructure, and cooperative views, as well as cooperative 3D tracking on the V2X-Seq benchmark; and 3) CRUISE effectively generates challenging corner cases. |
| title | CRUISE: Cooperative Reconstruction and Editing in V2X Scenarios using Gaussian Splatting |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2507.18473 |