Self-correcting online navigation via leveraged Gaussian processes

Seunggyu Chang; Sungjoon Choi; Songhwai Oh

doi:10.1109/URAI.2017.7992850

Self-correcting online navigation via leveraged Gaussian processes

Seunggyu Chang, Sungjoon Choi, Songhwai Oh

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

Abstract

In this paper, a novel online learning navigation algorithm is proposed to incorporate negative data generated from failure in an online manner. While existing methods require additional knowledge about what to do at failed situations, the proposed method alleviates this by utilizing failures as a clue of what not to do without requiring additional knowledge of what to do. By combining the benefits of leveraged Gaussian processes and sparse online Gaussian processes, we proposed an online learning framework for navigation and its update rule which instantly learns which actions to avoid from the failures while navigating. Our navigation method is successfully validated on a static planar world and dynamic worlds on both simulation and real-world dataset.

Original language	English
Title of host publication	2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	868-873
Number of pages	6
ISBN (Electronic)	9781509030552
DOIs	https://doi.org/10.1109/URAI.2017.7992850
Publication status	Published - 2017 Jul 25
Externally published	Yes
Event	14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017 - Jeju, Korea, Republic of Duration: 2017 Jun 28 → 2017 Jul 1

Publication series

Name	2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017

Other

Other	14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017
Country/Territory	Korea, Republic of
City	Jeju
Period	17/6/28 → 17/7/1

Bibliographical note

Funding Information:
ACKNOWLEDGMENT This work was supported by Institute for Information & communications Technology Promotion (IITP) grant funded by the Korea government (MSIP) (NO.2014-0-00147, Basic Software Research in Human-level Lifelong Machine Learning).

Publisher Copyright:
© 2017 IEEE.

Keywords

Learning from failure
Online learning navigation

ASJC Scopus subject areas

Computer Science Applications
Biomedical Engineering
Artificial Intelligence
Human-Computer Interaction
Control and Optimization

Access to Document

10.1109/URAI.2017.7992850

Cite this

Chang, S., Choi, S., & Oh, S. (2017). Self-correcting online navigation via leveraged Gaussian processes. In 2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017 (pp. 868-873). Article 7992850 (2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/URAI.2017.7992850

Self-correcting online navigation via leveraged Gaussian processes. / Chang, Seunggyu; Choi, Sungjoon; Oh, Songhwai.
2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017. Institute of Electrical and Electronics Engineers Inc., 2017. p. 868-873 7992850 (2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017).

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

Chang, S, Choi, S & Oh, S 2017, Self-correcting online navigation via leveraged Gaussian processes. in 2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017., 7992850, 2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017, Institute of Electrical and Electronics Engineers Inc., pp. 868-873, 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017, Jeju, Korea, Republic of, 17/6/28. https://doi.org/10.1109/URAI.2017.7992850

Chang S, Choi S, Oh S. Self-correcting online navigation via leveraged Gaussian processes. In 2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017. Institute of Electrical and Electronics Engineers Inc. 2017. p. 868-873. 7992850. (2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017). doi: 10.1109/URAI.2017.7992850

Chang, Seunggyu ; Choi, Sungjoon ; Oh, Songhwai. / Self-correcting online navigation via leveraged Gaussian processes. 2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017. Institute of Electrical and Electronics Engineers Inc., 2017. pp. 868-873 (2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017).

@inproceedings{6aa927cce57c40c38e63253f4e05980d,

title = "Self-correcting online navigation via leveraged Gaussian processes",

abstract = "In this paper, a novel online learning navigation algorithm is proposed to incorporate negative data generated from failure in an online manner. While existing methods require additional knowledge about what to do at failed situations, the proposed method alleviates this by utilizing failures as a clue of what not to do without requiring additional knowledge of what to do. By combining the benefits of leveraged Gaussian processes and sparse online Gaussian processes, we proposed an online learning framework for navigation and its update rule which instantly learns which actions to avoid from the failures while navigating. Our navigation method is successfully validated on a static planar world and dynamic worlds on both simulation and real-world dataset.",

keywords = "Learning from failure, Online learning navigation",

author = "Seunggyu Chang and Sungjoon Choi and Songhwai Oh",

note = "Funding Information: ACKNOWLEDGMENT This work was supported by Institute for Information & communications Technology Promotion (IITP) grant funded by the Korea government (MSIP) (NO.2014-0-00147, Basic Software Research in Human-level Lifelong Machine Learning). Publisher Copyright: {\textcopyright} 2017 IEEE.; 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017 ; Conference date: 28-06-2017 Through 01-07-2017",

year = "2017",

month = jul,

day = "25",

doi = "10.1109/URAI.2017.7992850",

language = "English",

series = "2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017",

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

pages = "868--873",

booktitle = "2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017",

}

TY - GEN

T1 - Self-correcting online navigation via leveraged Gaussian processes

AU - Chang, Seunggyu

AU - Choi, Sungjoon

AU - Oh, Songhwai

N1 - Funding Information: ACKNOWLEDGMENT This work was supported by Institute for Information & communications Technology Promotion (IITP) grant funded by the Korea government (MSIP) (NO.2014-0-00147, Basic Software Research in Human-level Lifelong Machine Learning). Publisher Copyright: © 2017 IEEE.

PY - 2017/7/25

Y1 - 2017/7/25

N2 - In this paper, a novel online learning navigation algorithm is proposed to incorporate negative data generated from failure in an online manner. While existing methods require additional knowledge about what to do at failed situations, the proposed method alleviates this by utilizing failures as a clue of what not to do without requiring additional knowledge of what to do. By combining the benefits of leveraged Gaussian processes and sparse online Gaussian processes, we proposed an online learning framework for navigation and its update rule which instantly learns which actions to avoid from the failures while navigating. Our navigation method is successfully validated on a static planar world and dynamic worlds on both simulation and real-world dataset.

AB - In this paper, a novel online learning navigation algorithm is proposed to incorporate negative data generated from failure in an online manner. While existing methods require additional knowledge about what to do at failed situations, the proposed method alleviates this by utilizing failures as a clue of what not to do without requiring additional knowledge of what to do. By combining the benefits of leveraged Gaussian processes and sparse online Gaussian processes, we proposed an online learning framework for navigation and its update rule which instantly learns which actions to avoid from the failures while navigating. Our navigation method is successfully validated on a static planar world and dynamic worlds on both simulation and real-world dataset.

KW - Learning from failure

KW - Online learning navigation

UR - http://www.scopus.com/inward/record.url?scp=85034272794&partnerID=8YFLogxK

U2 - 10.1109/URAI.2017.7992850

DO - 10.1109/URAI.2017.7992850

M3 - Conference contribution

AN - SCOPUS:85034272794

T3 - 2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017

SP - 868

EP - 873

BT - 2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 14th International Conference on Ubiquitous Robots and Ambient Intelligence, URAI 2017

Y2 - 28 June 2017 through 1 July 2017

ER -

Self-correcting online navigation via leveraged Gaussian processes

Abstract

Publication series

Other

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this