Self-Supervised Pre-training with Combined Datasets for 3D Perception in Autonomous Driving
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866916693783085056 |
|---|---|
| author | Wang, Shumin Yang, Zhuoran Wang, Lidian Tang, Zhipeng Li, Heng Pan, Lehan Zhang, Sha Peng, Jie Ji, Jianmin Zhang, Yanyong |
| author_facet | Wang, Shumin Yang, Zhuoran Wang, Lidian Tang, Zhipeng Li, Heng Pan, Lehan Zhang, Sha Peng, Jie Ji, Jianmin Zhang, Yanyong |
| contents | The significant achievements of pre-trained models leveraging large volumes of data in the field of NLP and 2D vision inspire us to explore the potential of extensive data pre-training for 3D perception in autonomous driving. Toward this goal, this paper proposes to utilize massive unlabeled data from heterogeneous datasets to pre-train 3D perception models. We introduce a self-supervised pre-training framework that learns effective 3D representations from scratch on unlabeled data, combined with a prompt adapter based domain adaptation strategy to reduce dataset bias. The approach significantly improves model performance on downstream tasks such as 3D object detection, BEV segmentation, 3D object tracking, and occupancy prediction, and shows steady performance increase as the training data volume scales up, demonstrating the potential of continually benefit 3D perception models for autonomous driving. We will release the source code to inspire further investigations in the community. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_12709 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Self-Supervised Pre-training with Combined Datasets for 3D Perception in Autonomous Driving Wang, Shumin Yang, Zhuoran Wang, Lidian Tang, Zhipeng Li, Heng Pan, Lehan Zhang, Sha Peng, Jie Ji, Jianmin Zhang, Yanyong Computer Vision and Pattern Recognition The significant achievements of pre-trained models leveraging large volumes of data in the field of NLP and 2D vision inspire us to explore the potential of extensive data pre-training for 3D perception in autonomous driving. Toward this goal, this paper proposes to utilize massive unlabeled data from heterogeneous datasets to pre-train 3D perception models. We introduce a self-supervised pre-training framework that learns effective 3D representations from scratch on unlabeled data, combined with a prompt adapter based domain adaptation strategy to reduce dataset bias. The approach significantly improves model performance on downstream tasks such as 3D object detection, BEV segmentation, 3D object tracking, and occupancy prediction, and shows steady performance increase as the training data volume scales up, demonstrating the potential of continually benefit 3D perception models for autonomous driving. We will release the source code to inspire further investigations in the community. |
| title | Self-Supervised Pre-training with Combined Datasets for 3D Perception in Autonomous Driving |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2504.12709 |