EUVIMEDEUVIMED entdecken
Crossref

SpaceFormer: Spatial Position Contextual Semantics Embedding for Multi-View 3D Object Detection

Jiaqi Zhao, Huanfeng Hu, Wen-Liang Du, Yong Zhou, Kunyang Sun, Rui Yao, Abdulmotaleb El Saddik

IEEE Transactions on Intelligent Transportation Systems · 2026-03 · 27 (3) · 3685-3698

DOI: 10.1109/tits.2025.3639680 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Intelligent Transportation Systems, 2026-03. Verfasst von Jiaqi Zhao und 6 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 49 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Double FirstClass Project of China University of Mining and Technology for Independent Innovation and Social Service, Six Talent Peaks Project in Jiangsu Province.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

3D object detection aims to accurately localize and recognize objects in 3D space. It serves as a fundamental task for reliable perception in intelligent transportation systems, enabling the monitoring of diverse traffic participants such as vehicles, pedestrians, cyclists, and public transport. Recently, transformer-based methods have gained significant attention in multi-view 3D object detection due to their strong global reasoning capabilities. However, their limited capacity to model spatial positional information hinders accurate object localization, especially in complex and large-scale scenes. To address this limitation, SpaceFormer is proposed as a novel transformer-based multi-view 3D object detector. Specifically, a Contextual Visual Prompts Learning strategy is proposed to enhance the perception of small and sparse traffic participants by incorporating contextual priors. To further suppress background interference, a Semantics-guided Depth Estimation method is proposed to refine depth representations using high-level semantic information. Furthermore, a Spatial Position Embedding mechanism is proposed to improve the spatial localization capability of the transformer by integrating geometric position and polar spatial embedding. Extensive experiments on the nuScenes benchmark demonstrate that SpaceFormer achieves state-of-the-art performance with 55.5% mAP and 62.9% NDS. These improvements indicate not only methodological advances but also practical benefits for intelligent transportation systems, enhancing safety, reliability, and efficiency in real-world deployments.

Förderung

  • National Natural Science Foundation of China
  • Double FirstClass Project of China University of Mining and Technology for Independent Innovation and Social Service
  • Six Talent Peaks Project in Jiangsu Province

Zitieren

APA

Zhao, J., Hu, H., Du, W., Zhou, Y., Sun, K., Yao, R., & Saddik, A. E. (2026). SpaceFormer: Spatial Position Contextual Semantics Embedding for Multi-View 3D Object Detection. IEEE Transactions on Intelligent Transportation Systems, 27(3), 3685-3698. https://doi.org/10.1109/tits.2025.3639680

Vancouver

Zhao J, Hu H, Du W, Zhou Y, Sun K, Yao R, et al. SpaceFormer: Spatial Position Contextual Semantics Embedding for Multi-View 3D Object Detection. IEEE Transactions on Intelligent Transportation Systems. 2026;27(3):3685-3698. doi:10.1109/tits.2025.3639680

BibTeX
@article{Zhao2026,
  author = {Zhao, Jiaqi and Hu, Huanfeng and Du, Wen-Liang and Zhou, Yong and Sun, Kunyang and Yao, Rui and Saddik, Abdulmotaleb El},
  title = {SpaceFormer: Spatial Position Contextual Semantics Embedding for Multi-View 3D Object Detection},
  journal = {IEEE Transactions on Intelligent Transportation Systems},
  year = {2026},
  volume = {27},
  number = {3},
  pages = {3685--3698},
  doi = {10.1109/tits.2025.3639680}
}
Kostenlos starten →