LiteVLoc: Map-Lite Visual Localization for Image Goal Navigation

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Jiao, Jianhao, He, Jinhao, Liu, Changkun, Aegidius, Sebastian, Hu, Xiangcheng, Braud, Tristan, Kanoulas, Dimitrios
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909356278153216
author Jiao, Jianhao
He, Jinhao
Liu, Changkun
Aegidius, Sebastian
Hu, Xiangcheng
Braud, Tristan
Kanoulas, Dimitrios
author_facet Jiao, Jianhao
He, Jinhao
Liu, Changkun
Aegidius, Sebastian
Hu, Xiangcheng
Braud, Tristan
Kanoulas, Dimitrios
contents This paper presents LiteVLoc, a hierarchical visual localization framework that uses a lightweight topo-metric map to represent the environment. The method consists of three sequential modules that estimate camera poses in a coarse-to-fine manner. Unlike mainstream approaches relying on detailed 3D representations, LiteVLoc reduces storage overhead by leveraging learning-based feature matching and geometric solvers for metric pose estimation. A novel dataset for the map-free relocalization task is also introduced. Extensive experiments including localization and navigation in both simulated and real-world scenarios have validate the system's performance and demonstrated its precision and efficiency for large-scale deployment. Code and data will be made publicly available.
format Preprint
id arxiv_https___arxiv_org_abs_2410_04419
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle LiteVLoc: Map-Lite Visual Localization for Image Goal Navigation
Jiao, Jianhao
He, Jinhao
Liu, Changkun
Aegidius, Sebastian
Hu, Xiangcheng
Braud, Tristan
Kanoulas, Dimitrios
Robotics
Computer Vision and Pattern Recognition
This paper presents LiteVLoc, a hierarchical visual localization framework that uses a lightweight topo-metric map to represent the environment. The method consists of three sequential modules that estimate camera poses in a coarse-to-fine manner. Unlike mainstream approaches relying on detailed 3D representations, LiteVLoc reduces storage overhead by leveraging learning-based feature matching and geometric solvers for metric pose estimation. A novel dataset for the map-free relocalization task is also introduced. Extensive experiments including localization and navigation in both simulated and real-world scenarios have validate the system's performance and demonstrated its precision and efficiency for large-scale deployment. Code and data will be made publicly available.
title LiteVLoc: Map-Lite Visual Localization for Image Goal Navigation
topic Robotics
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2410.04419