A Vision-Centric Approach for Static Map Element Annotation

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Zhang, Jiaxin, Chen, Shiyuan, Yin, Haoran, Mei, Ruohong, Liu, Xuan, Yang, Cong, Zhang, Qian, Sui, Wei
Format: Preprint
Veröffentlicht: 2023
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866913236047101952
author Zhang, Jiaxin
Chen, Shiyuan
Yin, Haoran
Mei, Ruohong
Liu, Xuan
Yang, Cong
Zhang, Qian
Sui, Wei
author_facet Zhang, Jiaxin
Chen, Shiyuan
Yin, Haoran
Mei, Ruohong
Liu, Xuan
Yang, Cong
Zhang, Qian
Sui, Wei
contents The recent development of online static map element (a.k.a. HD Map) construction algorithms has raised a vast demand for data with ground truth annotations. However, available public datasets currently cannot provide high-quality training data regarding consistency and accuracy. To this end, we present CAMA: a vision-centric approach for Consistent and Accurate Map Annotation. Without LiDAR inputs, our proposed framework can still generate high-quality 3D annotations of static map elements. Specifically, the annotation can achieve high reprojection accuracy across all surrounding cameras and is spatial-temporal consistent across the whole sequence. We apply our proposed framework to the popular nuScenes dataset to provide efficient and highly accurate annotations. Compared with the original nuScenes static map element, models trained with annotations from CAMA achieve lower reprojection errors (e.g., 4.73 vs. 8.03 pixels).
format Preprint
id arxiv_https___arxiv_org_abs_2309_11754
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle A Vision-Centric Approach for Static Map Element Annotation
Zhang, Jiaxin
Chen, Shiyuan
Yin, Haoran
Mei, Ruohong
Liu, Xuan
Yang, Cong
Zhang, Qian
Sui, Wei
Computer Vision and Pattern Recognition
The recent development of online static map element (a.k.a. HD Map) construction algorithms has raised a vast demand for data with ground truth annotations. However, available public datasets currently cannot provide high-quality training data regarding consistency and accuracy. To this end, we present CAMA: a vision-centric approach for Consistent and Accurate Map Annotation. Without LiDAR inputs, our proposed framework can still generate high-quality 3D annotations of static map elements. Specifically, the annotation can achieve high reprojection accuracy across all surrounding cameras and is spatial-temporal consistent across the whole sequence. We apply our proposed framework to the popular nuScenes dataset to provide efficient and highly accurate annotations. Compared with the original nuScenes static map element, models trained with annotations from CAMA achieve lower reprojection errors (e.g., 4.73 vs. 8.03 pixels).
title A Vision-Centric Approach for Static Map Element Annotation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2309.11754