RAS_AA_FinalManuscript.pdf (2.67 MB)
Robust view based navigation through view classification
conference contribution
posted on 2023-06-10, 05:43 authored by Amany Said AminAmany Said Amin, Efstathios KagioulisEfstathios Kagioulis, Norbert DomcsekNorbert Domcsek, Paul GrahamPaul Graham, Thomas NowotnyThomas Nowotny, Andy PhilippidesAndy PhilippidesCurrent implementations of view-based navigation on robots have shown success, but are limited to routes of <10m [1] [2]. This is in part because current strategies do not take into account whether a view has been correctly recognised, moving in the most familiar direction given by the rotational familiarity function (RFF) regardless of prediction confidence. We demonstrate that it is possible to use the shape of the RFF to classify if the current view is from a known position, and thus likely to provide valid navigational information, or from a position which is unknown, aliased or occluded and therefore likely to result in erroneous movement. Our model could classify these four view types with accuracies of 1.00, 0.91, 0.97 and 0.87 respectively. We hope to use these results to extend online view-based navigation and prevent robot loss in complex environments.
History
Publication status
- Published
File Version
- Accepted version
Journal
UKRAS22 Conference "Robotics for Unconstrained Environments" ProceedingsISSN
2516-502XPublisher
EPSRC UK-RAS NetworkExternal DOI
Volume
5Page range
76-77Event name
UKRAS22 Conference "Robotics for Unconstrained Environments"Event location
Aberystwyth University, WalesEvent type
conferenceEvent date
26 August 2022Department affiliated with
- Informatics Publications
Full text available
- Yes
Legacy Posted Date
2022-12-20First Open Access (FOA) Date
2022-12-20First Compliant Deposit (FCD) Date
2022-12-19Usage metrics
Categories
No categories selectedKeywords
Licence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC