3DSFLabelling: Boosting 3D Scene Flow Estimation by Pseudo Auto-labelling
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916143615180800 |
|---|---|
| author | Jiang, Chaokang Wang, Guangming Liu, Jiuming Wang, Hesheng Ma, Zhuang Liu, Zhenqiang Liang, Zhujin Shan, Yi Du, Dalong |
| author_facet | Jiang, Chaokang Wang, Guangming Liu, Jiuming Wang, Hesheng Ma, Zhuang Liu, Zhenqiang Liang, Zhujin Shan, Yi Du, Dalong |
| contents | Learning 3D scene flow from LiDAR point clouds presents significant difficulties, including poor generalization from synthetic datasets to real scenes, scarcity of real-world 3D labels, and poor performance on real sparse LiDAR point clouds. We present a novel approach from the perspective of auto-labelling, aiming to generate a large number of 3D scene flow pseudo labels for real-world LiDAR point clouds. Specifically, we employ the assumption of rigid body motion to simulate potential object-level rigid movements in autonomous driving scenarios. By updating different motion attributes for multiple anchor boxes, the rigid motion decomposition is obtained for the whole scene. Furthermore, we developed a novel 3D scene flow data augmentation method for global and local motion. By perfectly synthesizing target point clouds based on augmented motion parameters, we easily obtain lots of 3D scene flow labels in point clouds highly consistent with real scenarios. On multiple real-world datasets including LiDAR KITTI, nuScenes, and Argoverse, our method outperforms all previous supervised and unsupervised methods without requiring manual labelling. Impressively, our method achieves a tenfold reduction in EPE3D metric on the LiDAR KITTI dataset, reducing it from $0.190m$ to a mere $0.008m$ error. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_18146 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | 3DSFLabelling: Boosting 3D Scene Flow Estimation by Pseudo Auto-labelling Jiang, Chaokang Wang, Guangming Liu, Jiuming Wang, Hesheng Ma, Zhuang Liu, Zhenqiang Liang, Zhujin Shan, Yi Du, Dalong Computer Vision and Pattern Recognition Learning 3D scene flow from LiDAR point clouds presents significant difficulties, including poor generalization from synthetic datasets to real scenes, scarcity of real-world 3D labels, and poor performance on real sparse LiDAR point clouds. We present a novel approach from the perspective of auto-labelling, aiming to generate a large number of 3D scene flow pseudo labels for real-world LiDAR point clouds. Specifically, we employ the assumption of rigid body motion to simulate potential object-level rigid movements in autonomous driving scenarios. By updating different motion attributes for multiple anchor boxes, the rigid motion decomposition is obtained for the whole scene. Furthermore, we developed a novel 3D scene flow data augmentation method for global and local motion. By perfectly synthesizing target point clouds based on augmented motion parameters, we easily obtain lots of 3D scene flow labels in point clouds highly consistent with real scenarios. On multiple real-world datasets including LiDAR KITTI, nuScenes, and Argoverse, our method outperforms all previous supervised and unsupervised methods without requiring manual labelling. Impressively, our method achieves a tenfold reduction in EPE3D metric on the LiDAR KITTI dataset, reducing it from $0.190m$ to a mere $0.008m$ error. |
| title | 3DSFLabelling: Boosting 3D Scene Flow Estimation by Pseudo Auto-labelling |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2402.18146 |