Visual feature selection for GP-based localization using an omnidirectional camera

Huan N. Do; Jongeun Choi; Chae Young Lim

doi:10.1109/ACC.2015.7171990

Visual feature selection for GP-based localization using an omnidirectional camera

Huan N. Do, Jongeun Choi, Chae Young Lim

Department of Mechanical Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

This paper considers visual feature selection and its regression to estimate the position of a vehicle using an omnidirectional camera. The Gaussian process (GP)-based localization builds on a maximum likelihood estimation (MLE) with a GP regression from optimally selected visual features. In particular, the collection of selected features over a surveillance region is modeled by a multivariate GP with unknown hyperparameters. The hyperparameters are identified through the learning process as the corresponding MLEs and they are used for prediction in an empirical Bayes fashion. To select features, we apply a backward sequential elimination technique in order to improve the quality of the position estimation with reduced number of features for efficient GP-based localization. The excellent results of the proposed algorithm from the real-world outdoor experimental study are illustrated using different visual features.

Original language	English
Title of host publication	ACC 2015 - 2015 American Control Conference
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	4210-4215
Number of pages	6
ISBN (Electronic)	9781479986842
DOIs	https://doi.org/10.1109/ACC.2015.7171990
Publication status	Published - 2015 Jul 28
Event	2015 American Control Conference, ACC 2015 - Chicago, United States Duration: 2015 Jul 1 → 2015 Jul 3

Publication series

Name	Proceedings of the American Control Conference
Volume	2015-July
ISSN (Print)	0743-1619

Other

Other	2015 American Control Conference, ACC 2015
Country/Territory	United States
City	Chicago
Period	15/7/1 → 15/7/3

Bibliographical note

Publisher Copyright:
© 2015 American Automatic Control Council.

All Science Journal Classification (ASJC) codes

Electrical and Electronic Engineering

Access to Document

10.1109/ACC.2015.7171990

Cite this

@inproceedings{dd5a4e31c23c49daad8b2cde4b994ae3,

title = "Visual feature selection for GP-based localization using an omnidirectional camera",

abstract = "This paper considers visual feature selection and its regression to estimate the position of a vehicle using an omnidirectional camera. The Gaussian process (GP)-based localization builds on a maximum likelihood estimation (MLE) with a GP regression from optimally selected visual features. In particular, the collection of selected features over a surveillance region is modeled by a multivariate GP with unknown hyperparameters. The hyperparameters are identified through the learning process as the corresponding MLEs and they are used for prediction in an empirical Bayes fashion. To select features, we apply a backward sequential elimination technique in order to improve the quality of the position estimation with reduced number of features for efficient GP-based localization. The excellent results of the proposed algorithm from the real-world outdoor experimental study are illustrated using different visual features.",

author = "Do, {Huan N.} and Jongeun Choi and Lim, {Chae Young}",

note = "Publisher Copyright: {\textcopyright} 2015 American Automatic Control Council.; 2015 American Control Conference, ACC 2015 ; Conference date: 01-07-2015 Through 03-07-2015",

year = "2015",

month = jul,

day = "28",

doi = "10.1109/ACC.2015.7171990",

language = "English",

series = "Proceedings of the American Control Conference",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "4210--4215",

booktitle = "ACC 2015 - 2015 American Control Conference",

address = "United States",

}

Do, HN, Choi, J & Lim, CY 2015, Visual feature selection for GP-based localization using an omnidirectional camera. in ACC 2015 - 2015 American Control Conference., 7171990, Proceedings of the American Control Conference, vol. 2015-July, Institute of Electrical and Electronics Engineers Inc., pp. 4210-4215, 2015 American Control Conference, ACC 2015, Chicago, United States, 15/7/1. https://doi.org/10.1109/ACC.2015.7171990

Visual feature selection for GP-based localization using an omnidirectional camera. / Do, Huan N.; Choi, Jongeun; Lim, Chae Young.
ACC 2015 - 2015 American Control Conference. Institute of Electrical and Electronics Engineers Inc., 2015. p. 4210-4215 7171990 (Proceedings of the American Control Conference; Vol. 2015-July).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Visual feature selection for GP-based localization using an omnidirectional camera

AU - Do, Huan N.

AU - Choi, Jongeun

AU - Lim, Chae Young

PY - 2015/7/28

Y1 - 2015/7/28

N2 - This paper considers visual feature selection and its regression to estimate the position of a vehicle using an omnidirectional camera. The Gaussian process (GP)-based localization builds on a maximum likelihood estimation (MLE) with a GP regression from optimally selected visual features. In particular, the collection of selected features over a surveillance region is modeled by a multivariate GP with unknown hyperparameters. The hyperparameters are identified through the learning process as the corresponding MLEs and they are used for prediction in an empirical Bayes fashion. To select features, we apply a backward sequential elimination technique in order to improve the quality of the position estimation with reduced number of features for efficient GP-based localization. The excellent results of the proposed algorithm from the real-world outdoor experimental study are illustrated using different visual features.

AB - This paper considers visual feature selection and its regression to estimate the position of a vehicle using an omnidirectional camera. The Gaussian process (GP)-based localization builds on a maximum likelihood estimation (MLE) with a GP regression from optimally selected visual features. In particular, the collection of selected features over a surveillance region is modeled by a multivariate GP with unknown hyperparameters. The hyperparameters are identified through the learning process as the corresponding MLEs and they are used for prediction in an empirical Bayes fashion. To select features, we apply a backward sequential elimination technique in order to improve the quality of the position estimation with reduced number of features for efficient GP-based localization. The excellent results of the proposed algorithm from the real-world outdoor experimental study are illustrated using different visual features.

UR - http://www.scopus.com/inward/record.url?scp=84940916950&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84940916950&partnerID=8YFLogxK

U2 - 10.1109/ACC.2015.7171990

DO - 10.1109/ACC.2015.7171990

M3 - Conference contribution

AN - SCOPUS:84940916950

T3 - Proceedings of the American Control Conference

SP - 4210

EP - 4215

BT - ACC 2015 - 2015 American Control Conference

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2015 American Control Conference, ACC 2015

Y2 - 1 July 2015 through 3 July 2015

ER -

Visual feature selection for GP-based localization using an omnidirectional camera

Abstract

Publication series

Other

Bibliographical note

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this