UnIRe: Unsupervised Instance Decomposition for Dynamic Urban Scene Reconstruction
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914037347909632 |
|---|---|
| author | Mao, Yunxuan Xiong, Rong Wang, Yue Liao, Yiyi |
| author_facet | Mao, Yunxuan Xiong, Rong Wang, Yue Liao, Yiyi |
| contents | Reconstructing and decomposing dynamic urban scenes is crucial for autonomous driving, urban planning, and scene editing. However, existing methods fail to perform instance-aware decomposition without manual annotations, which is crucial for instance-level scene editing.We propose UnIRe, a 3D Gaussian Splatting (3DGS) based approach that decomposes a scene into a static background and individual dynamic instances using only RGB images and LiDAR point clouds. At its core, we introduce 4D superpoints, a novel representation that clusters multi-frame LiDAR points in 4D space, enabling unsupervised instance separation based on spatiotemporal correlations. These 4D superpoints serve as the foundation for our decomposed 4D initialization, i.e., providing spatial and temporal initialization to train a dynamic 3DGS for arbitrary dynamic classes without requiring bounding boxes or object templates.Furthermore, we introduce a smoothness regularization strategy in both 2D and 3D space, further improving the temporal stability.Experiments on benchmark datasets show that our method outperforms existing methods in decomposed dynamic scene reconstruction while enabling accurate and flexible instance-level editing, making it a practical solution for real-world applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_00763 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | UnIRe: Unsupervised Instance Decomposition for Dynamic Urban Scene Reconstruction Mao, Yunxuan Xiong, Rong Wang, Yue Liao, Yiyi Computer Vision and Pattern Recognition Robotics Reconstructing and decomposing dynamic urban scenes is crucial for autonomous driving, urban planning, and scene editing. However, existing methods fail to perform instance-aware decomposition without manual annotations, which is crucial for instance-level scene editing.We propose UnIRe, a 3D Gaussian Splatting (3DGS) based approach that decomposes a scene into a static background and individual dynamic instances using only RGB images and LiDAR point clouds. At its core, we introduce 4D superpoints, a novel representation that clusters multi-frame LiDAR points in 4D space, enabling unsupervised instance separation based on spatiotemporal correlations. These 4D superpoints serve as the foundation for our decomposed 4D initialization, i.e., providing spatial and temporal initialization to train a dynamic 3DGS for arbitrary dynamic classes without requiring bounding boxes or object templates.Furthermore, we introduce a smoothness regularization strategy in both 2D and 3D space, further improving the temporal stability.Experiments on benchmark datasets show that our method outperforms existing methods in decomposed dynamic scene reconstruction while enabling accurate and flexible instance-level editing, making it a practical solution for real-world applications. |
| title | UnIRe: Unsupervised Instance Decomposition for Dynamic Urban Scene Reconstruction |
| topic | Computer Vision and Pattern Recognition Robotics |
| url | https://arxiv.org/abs/2504.00763 |