Vision-based Uneven BEV Representation Learning with Polar Rasterization and Surface Estimation
Zhi Liu, Shaoyu Chen, Xiaojie Guo, Xinggang Wang, Tianheng Cheng, Hongmei Zhu, Qian Zhang, Wenyu Liu
Abstract
In this work, we propose PolarBEV for vision-based uneven BEV representation learning. To adapt to the foreshortening effect of camera imaging, we rasterize the BEV space both angularly and radially, and introduce polar embedding decomposition to model the associations among polar grids. Polar grids are rearranged to an array-like regular representation for efficient processing. Besides, to determine the 2D-to-3D correspondence, we iteratively update the BEV surface based on a hypothetical plane, and adopt height-based feature transformation. PolarBEV keeps real-time inference speed on a single 2080Ti GPU, and outperforms other methods for both BEV semantic segmentation and BEV instance segmentation. Thorough ablations are presented to validate the design. The code will be released for facilitating further research.
BibTeX
@inproceedings{
liu2022visionbased,
title={Vision-based Uneven {BEV} Representation Learning with Polar Rasterization and Surface Estimation},
author={Zhi Liu and Shaoyu Chen and Xiaojie Guo and Xinggang Wang and Tianheng Cheng and Hongmei Zhu and Qian Zhang and Wenyu Liu and Yi Zhang},
booktitle={6th Annual Conference on Robot Learning},
year={2022},
url={https://openreview.net/forum?id=SM70KHTBG-0}
}