Generalized Proximal Methods for Pose Graph Optimization

Taosha Fan; Todd Murphey

doi:10.1007/978-3-030-95459-8_24

Generalized Proximal Methods for Pose Graph Optimization

Taosha Fan^*, Todd Murphey

^*Corresponding author for this work

Mechanical Engineering

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

2 Scopus citations

Abstract

In this paper, we generalize proximal methods that were originally designed for convex optimization on normed vector space to non-convex pose graph optimization (PGO) on special Euclidean groups, and show that our proposed generalized proximal methods for PGO converge to first-order critical points. Furthermore, we propose methods that significantly accelerate the rates of convergence almost without loss of any theoretical guarantees. In addition, our proposed methods can be easily distributed and parallelized with no compromise of efficiency. The efficacy of this work is validated through implementation on simultaneous localization and mapping (SLAM) and distributed 3D sensor network localization, which indicate that our proposed methods are a lot faster than existing techniques to converge to sufficient accuracy for practical use.

Original language	English (US)
Title of host publication	Robotics Research - The 19th International Symposium ISRR
Editors	Tamim Asfour, Eiichi Yoshida, Jaeheung Park, Henrik Christensen, Oussama Khatib
Publisher	Springer Nature
Pages	393-409
Number of pages	17
ISBN (Print)	9783030954581
DOIs	https://doi.org/10.1007/978-3-030-95459-8_24
State	Published - 2022
Event	17th International Symposium of Robotics Research, ISRR 2019 - Hanoi, Viet Nam Duration: Oct 6 2019 → Oct 10 2019

Publication series

Name	Springer Proceedings in Advanced Robotics
Volume	20 SPAR
ISSN (Print)	2511-1256
ISSN (Electronic)	2511-1264

Conference

Conference	17th International Symposium of Robotics Research, ISRR 2019
Country/Territory	Viet Nam
City	Hanoi
Period	10/6/19 → 10/10/19

ASJC Scopus subject areas

Control and Systems Engineering
Electrical and Electronic Engineering
Mechanical Engineering
Engineering (miscellaneous)
Artificial Intelligence
Computer Science Applications
Applied Mathematics

Access to Document

10.1007/978-3-030-95459-8_24

Cite this

@inproceedings{9b7f382f15fe49cf9c6b69fa57036459,

title = "Generalized Proximal Methods for Pose Graph Optimization",

abstract = "In this paper, we generalize proximal methods that were originally designed for convex optimization on normed vector space to non-convex pose graph optimization (PGO) on special Euclidean groups, and show that our proposed generalized proximal methods for PGO converge to first-order critical points. Furthermore, we propose methods that significantly accelerate the rates of convergence almost without loss of any theoretical guarantees. In addition, our proposed methods can be easily distributed and parallelized with no compromise of efficiency. The efficacy of this work is validated through implementation on simultaneous localization and mapping (SLAM) and distributed 3D sensor network localization, which indicate that our proposed methods are a lot faster than existing techniques to converge to sufficient accuracy for practical use.",

author = "Taosha Fan and Todd Murphey",

note = "Publisher Copyright: {\textcopyright} 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.; 17th International Symposium of Robotics Research, ISRR 2019 ; Conference date: 06-10-2019 Through 10-10-2019",

year = "2022",

doi = "10.1007/978-3-030-95459-8_24",

language = "English (US)",

isbn = "9783030954581",

series = "Springer Proceedings in Advanced Robotics",

publisher = "Springer Nature",

pages = "393--409",

editor = "Tamim Asfour and Eiichi Yoshida and Jaeheung Park and Henrik Christensen and Oussama Khatib",

booktitle = "Robotics Research - The 19th International Symposium ISRR",

}

Fan, T & Murphey, T 2022, Generalized Proximal Methods for Pose Graph Optimization. in T Asfour, E Yoshida, J Park, H Christensen & O Khatib (eds), Robotics Research - The 19th International Symposium ISRR. Springer Proceedings in Advanced Robotics, vol. 20 SPAR, Springer Nature, pp. 393-409, 17th International Symposium of Robotics Research, ISRR 2019, Hanoi, Viet Nam, 10/6/19. https://doi.org/10.1007/978-3-030-95459-8_24

Generalized Proximal Methods for Pose Graph Optimization. / Fan, Taosha; Murphey, Todd.
Robotics Research - The 19th International Symposium ISRR. ed. / Tamim Asfour; Eiichi Yoshida; Jaeheung Park; Henrik Christensen; Oussama Khatib. Springer Nature, 2022. p. 393-409 (Springer Proceedings in Advanced Robotics; Vol. 20 SPAR).

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

TY - GEN

T1 - Generalized Proximal Methods for Pose Graph Optimization

AU - Fan, Taosha

AU - Murphey, Todd

PY - 2022

Y1 - 2022

N2 - In this paper, we generalize proximal methods that were originally designed for convex optimization on normed vector space to non-convex pose graph optimization (PGO) on special Euclidean groups, and show that our proposed generalized proximal methods for PGO converge to first-order critical points. Furthermore, we propose methods that significantly accelerate the rates of convergence almost without loss of any theoretical guarantees. In addition, our proposed methods can be easily distributed and parallelized with no compromise of efficiency. The efficacy of this work is validated through implementation on simultaneous localization and mapping (SLAM) and distributed 3D sensor network localization, which indicate that our proposed methods are a lot faster than existing techniques to converge to sufficient accuracy for practical use.

AB - In this paper, we generalize proximal methods that were originally designed for convex optimization on normed vector space to non-convex pose graph optimization (PGO) on special Euclidean groups, and show that our proposed generalized proximal methods for PGO converge to first-order critical points. Furthermore, we propose methods that significantly accelerate the rates of convergence almost without loss of any theoretical guarantees. In addition, our proposed methods can be easily distributed and parallelized with no compromise of efficiency. The efficacy of this work is validated through implementation on simultaneous localization and mapping (SLAM) and distributed 3D sensor network localization, which indicate that our proposed methods are a lot faster than existing techniques to converge to sufficient accuracy for practical use.

UR - http://www.scopus.com/inward/record.url?scp=85126248782&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85126248782&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-95459-8_24

DO - 10.1007/978-3-030-95459-8_24

M3 - Conference contribution

AN - SCOPUS:85126248782

SN - 9783030954581

T3 - Springer Proceedings in Advanced Robotics

SP - 393

EP - 409

BT - Robotics Research - The 19th International Symposium ISRR

A2 - Asfour, Tamim

A2 - Yoshida, Eiichi

A2 - Park, Jaeheung

A2 - Christensen, Henrik

A2 - Khatib, Oussama

PB - Springer Nature

T2 - 17th International Symposium of Robotics Research, ISRR 2019

Y2 - 6 October 2019 through 10 October 2019

ER -