/DA-BEV

DA-BEV: Unsupervised Domain Adaptation for Bird's Eye View Perception

DA-BEV: Unsupervised Domain Adaptation for Bird's Eye View Perception

DA-BEV: Unsupervised Domain Adaptation for Bird's Eye View Perception

Abstract

Camera-only Bird's Eye View (BEV) has demonstrated great potential in environment perception in a 3D space. However, most existing studies were conducted under a supervised setup which cannot scale well while handling various new data. Unsupervised domain adaptive BEV, which effective learning from various unlabelled target data, is far under-explored. In this work, we design DA-BEV, the first domain adaptive camera-only BEV framework that addresses domain adaptive BEV challenges by exploiting the complementary nature of image-view features and BEV features. DA-BEV introduces the idea of query into the domain adaptation framework to derive useful information from image-view and BEV features. It consists of two query-based designs, namely, query-based adversarial learning (QAL) and query-based self-training (QST), which exploits image-view features or BEV features to regularize the adaptation of the other. Extensive experiments show that DA-BEV achieves superior domain adaptive BEV perception performance consistently across multiple datasets and tasks such as 3D object detection and 3D scene segmentation.

Citation

@inproceedings{jiang2025bev,
  title={Da-bev: Unsupervised domain adaptation for bird’s eye view perception},
  author={Jiang, Kai and Huang, Jiaxing and Xie, Weiying and Lei, Jie and Li, Yunsong and Shao, Ling and Lu, Shijian},
  booktitle={European Conference on Computer Vision},
  pages={322--341},
  year={2025},
  organization={Springer}
}