arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2207.01878cs.CV
收藏

基于视觉的非均匀BEV极坐标光栅化和曲面估计的表示学习

Vision-based Uneven BEV Representation Learning with Polar Rasterization and Surface Estimation

Tianjin University(天津大学);Horizon Robotics(地平线机器人);Huazhong University of Science & Technology(华中科技大学)

Zhi Liu, Shaoyu Chen, Xiaojie Guo, Xinggang Wang, Tianheng Cheng, Hongmei Zhu, Qian Zhang, Wenyu Liu, Yi Zhang

摘要:

在这项工作中,我们提出了基于视觉的不均匀BEV表示学习的PolarBEV。为了适应相机成像的透视收缩效应,我们对BEV空间进行了角向和径向的栅格化,并引入极坐标嵌入分解来模拟极坐标网格之间的关联。极坐标网格被重新排列为类似于阵列的规则表示,以提高处理效率。此外,为了确定2D到3D的对应关系,我们基于假设平面迭代更新BEV表面,并采用基于高度的特征变换。PolarBEV在单个2080 Ti GPU上保持实时推理速度,并在BEV语义分割和BEV实例分割方面优于其他方法。提供彻底消融以验证设计。代码将在\url{https://github.com/SuperZ-Liu/PolarBEV}发布。

英文摘要:

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 at \url{https://github.com/SuperZ-Liu/PolarBEV}.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑