EUVIMEDEUVIMED entdecken
Crossref

DriveFlow: Rectified Flow Adaptation for Robust 3D Object Detection in Autonomous Driving

Hongbin Lin, Yiming Yang, Chaoda Zheng, Yifan Zhang, Shuaicheng Niu, Zilu Guo, Yafeng Li, Gui Gui, Shuguang Cui, Zhen Li

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (9) · 6943-6951

DOI: 10.1609/aaai.v40i9.37628 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Hongbin Lin und 9 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In autonomous driving, vision-centric 3D object detection recognizes and localizes 3D objects from RGB images. However, due to high annotation costs and diverse outdoor scenes, training data often fails to cover all possible test scenarios, known as the out-of-distribution (OOD) issue. Training-free image editing offers a promising solution for improving model robustness by training data enhancement without any modifications to pre-trained diffusion models. Nevertheless, inversion-based methods often suffer from limited effectiveness and inherent inaccuracies, while recent rectified-flow-based approaches struggle to preserve objects with accurate 3D geometry. In this paper, we propose DriveFlow, a Rectified Flow Adaptation method for training data enhancement in autonomous driving based on pre-trained Text-to-Image flow models. Based on frequency decomposition, DriveFlow introduces two strategies to adapt noise-free editing paths derived from text-conditioned velocities. 1) High-Frequency Foreground Preservation: DriveFlow incorporates a high-frequency alignment loss for foreground to maintain precise 3D object geometry. 2) Dual-Frequency Background Optimization: DriveFlow also conducts dual-frequency optimization for background, balancing editing flexibility and semantic consistency. Comprehensive experiments validate the effectiveness and efficiency of DriveFlow, demonstrating comprehensive performance improvements on all categories across OOD scenarios.

Zitieren

APA

Lin, H., Yang, Y., Zheng, C., Zhang, Y., Niu, S., Guo, Z., Li, Y., Gui, G., Cui, S., & Li, Z. (2026). DriveFlow: Rectified Flow Adaptation for Robust 3D Object Detection in Autonomous Driving. Proceedings of the AAAI Conference on Artificial Intelligence, 40(9), 6943-6951. https://doi.org/10.1609/aaai.v40i9.37628

Vancouver

Lin H, Yang Y, Zheng C, Zhang Y, Niu S, Guo Z, et al. DriveFlow: Rectified Flow Adaptation for Robust 3D Object Detection in Autonomous Driving. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(9):6943-6951. doi:10.1609/aaai.v40i9.37628

BibTeX
@article{Lin2026,
  author = {Lin, Hongbin and Yang, Yiming and Zheng, Chaoda and Zhang, Yifan and Niu, Shuaicheng and Guo, Zilu and Li, Yafeng and Gui, Gui and Cui, Shuguang and Li, Zhen},
  title = {DriveFlow: Rectified Flow Adaptation for Robust 3D Object Detection in Autonomous Driving},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {9},
  pages = {6943--6951},
  doi = {10.1609/aaai.v40i9.37628}
}
Kostenlos starten →