Predictive World Models for Social Navigation

Goodluck Oguzie; Anikó Ekárt; Luis J. Manso

doi:10.1007/978-3-031-47508-5_5

Predictive World Models for Social Navigation

Goodluck Oguzie, Anikó Ekárt, Luis J. Manso

Research output: Chapter in Book/Published conference output › Conference publication

Abstract

As robots begin to coexist with humans, the need for efficient and safe social robot navigation becomes increasingly pressing. In this paper we investigate how world models can enhance the effectiveness of reinforcement learning in social navigation tasks. We introduce three approaches that leverage predictive world models, which are then benchmarked against state-of-the-art algorithms. For a comprehensive and reliable evaluation, we employed multiple metrics during the training and testing phases. The key novelty of our approach consists in the integration and evaluation of predictive world models within the context of social navigation, as well as in the models themselves. Based on a diverse set of performance metrics, the experimental results provide evidence that predictive world models help improve reinforcement learning techniques for social navigation.

Original language	English
Title of host publication	Advances in Computational Intelligence Systems, Contributions Presented at the 22nd UK Workshop on Computational Intelligence
Editors	Nitin Naik, Paul Jenkins, Paul Grace, Longzhi Yang, Shaligram Prajapat
Publisher	Springer
Pages	53–64
Number of pages	13
ISBN (Electronic)	978-3-031-47508-5
ISBN (Print)	978-3-031-47507-8
DOIs	https://doi.org/10.1007/978-3-031-47508-5_5
Publication status	Published - 1 Feb 2024
Event	The 22nd UK Workship on Computational Intelligence - Aston University, Birmingham, United Kingdom Duration: 6 Sept 2023 → 8 Sept 2023 https://www.uk-ci.org/home

Publication series

Name	Advances in Intelligent Systems and Computing (AISC)
Publisher	Springer
ISSN (Print)	2194-5357
ISSN (Electronic)	2194-5365

Conference

Conference	The 22nd UK Workship on Computational Intelligence
Abbreviated title	UKCI 2023
Country/Territory	United Kingdom
City	Birmingham
Period	6/09/23 → 8/09/23
Internet address	https://www.uk-ci.org/home

Bibliographical note

This version of the paper has been accepted for publication, after peer review and is subject to Springer Nature’s AM terms of use [https://www.springernature.com/gp/open-research/policies/accepted-manuscript-terms], but is not the Version of Record and does not reflect post-acceptance improvements, or any corrections. The Version of Record is available online at: https://doi.org/10.1007/978-3-031-47508-5_5

Access to Document

10.1007/978-3-031-47508-5_5

Cite this

Oguzie, G., Ekárt, A., & Manso, L. J. (2024). Predictive World Models for Social Navigation. In N. Naik, P. Jenkins, P. Grace, L. Yang, & S. Prajapat (Eds.), Advances in Computational Intelligence Systems, Contributions Presented at the 22nd UK Workshop on Computational Intelligence (pp. 53–64). (Advances in Intelligent Systems and Computing (AISC)). Springer. https://doi.org/10.1007/978-3-031-47508-5_5

Oguzie, Goodluck ; Ekárt, Anikó ; Manso, Luis J. / Predictive World Models for Social Navigation. Advances in Computational Intelligence Systems, Contributions Presented at the 22nd UK Workshop on Computational Intelligence. editor / Nitin Naik ; Paul Jenkins ; Paul Grace ; Longzhi Yang ; Shaligram Prajapat. Springer, 2024. pp. 53–64 (Advances in Intelligent Systems and Computing (AISC)).

@inproceedings{6ce46683ecc84a399dd3d76d0611d7bf,

title = "Predictive World Models for Social Navigation",

abstract = "As robots begin to coexist with humans, the need for efficient and safe social robot navigation becomes increasingly pressing. In this paper we investigate how world models can enhance the effectiveness of reinforcement learning in social navigation tasks. We introduce three approaches that leverage predictive world models, which are then benchmarked against state-of-the-art algorithms. For a comprehensive and reliable evaluation, we employed multiple metrics during the training and testing phases. The key novelty of our approach consists in the integration and evaluation of predictive world models within the context of social navigation, as well as in the models themselves. Based on a diverse set of performance metrics, the experimental results provide evidence that predictive world models help improve reinforcement learning techniques for social navigation.",

author = "Goodluck Oguzie and Anik{\'o} Ek{\'a}rt and Manso, {Luis J.}",

note = "This version of the paper has been accepted for publication, after peer review and is subject to Springer Nature{\textquoteright}s AM terms of use [https://www.springernature.com/gp/open-research/policies/accepted-manuscript-terms], but is not the Version of Record and does not reflect post-acceptance improvements, or any corrections. The Version of Record is available online at: https://doi.org/10.1007/978-3-031-47508-5_5; The 22nd UK Workship on Computational Intelligence, UKCI 2023 ; Conference date: 06-09-2023 Through 08-09-2023",

year = "2024",

month = feb,

day = "1",

doi = "10.1007/978-3-031-47508-5_5",

language = "English",

isbn = "978-3-031-47507-8",

series = "Advances in Intelligent Systems and Computing (AISC)",

publisher = "Springer",

pages = "53–64",

editor = "Nitin Naik and Paul Jenkins and Paul Grace and Longzhi Yang and Shaligram Prajapat",

booktitle = "Advances in Computational Intelligence Systems, Contributions Presented at the 22nd UK Workshop on Computational Intelligence",

address = "Germany",

url = "https://www.uk-ci.org/home",

}

Oguzie, G, Ekárt, A & Manso, LJ 2024, Predictive World Models for Social Navigation. in N Naik, P Jenkins, P Grace, L Yang & S Prajapat (eds), Advances in Computational Intelligence Systems, Contributions Presented at the 22nd UK Workshop on Computational Intelligence. Advances in Intelligent Systems and Computing (AISC), Springer, pp. 53–64, The 22nd UK Workship on Computational Intelligence, Birmingham, United Kingdom, 6/09/23. https://doi.org/10.1007/978-3-031-47508-5_5

Predictive World Models for Social Navigation. / Oguzie, Goodluck; Ekárt, Anikó ; Manso, Luis J.
Advances in Computational Intelligence Systems, Contributions Presented at the 22nd UK Workshop on Computational Intelligence. ed. / Nitin Naik; Paul Jenkins; Paul Grace; Longzhi Yang; Shaligram Prajapat. Springer, 2024. p. 53–64 (Advances in Intelligent Systems and Computing (AISC)).

Research output: Chapter in Book/Published conference output › Conference publication

TY - GEN

T1 - Predictive World Models for Social Navigation

AU - Oguzie, Goodluck

AU - Ekárt, Anikó

AU - Manso, Luis J.

N1 - This version of the paper has been accepted for publication, after peer review and is subject to Springer Nature’s AM terms of use [https://www.springernature.com/gp/open-research/policies/accepted-manuscript-terms], but is not the Version of Record and does not reflect post-acceptance improvements, or any corrections. The Version of Record is available online at: https://doi.org/10.1007/978-3-031-47508-5_5

PY - 2024/2/1

Y1 - 2024/2/1

N2 - As robots begin to coexist with humans, the need for efficient and safe social robot navigation becomes increasingly pressing. In this paper we investigate how world models can enhance the effectiveness of reinforcement learning in social navigation tasks. We introduce three approaches that leverage predictive world models, which are then benchmarked against state-of-the-art algorithms. For a comprehensive and reliable evaluation, we employed multiple metrics during the training and testing phases. The key novelty of our approach consists in the integration and evaluation of predictive world models within the context of social navigation, as well as in the models themselves. Based on a diverse set of performance metrics, the experimental results provide evidence that predictive world models help improve reinforcement learning techniques for social navigation.

AB - As robots begin to coexist with humans, the need for efficient and safe social robot navigation becomes increasingly pressing. In this paper we investigate how world models can enhance the effectiveness of reinforcement learning in social navigation tasks. We introduce three approaches that leverage predictive world models, which are then benchmarked against state-of-the-art algorithms. For a comprehensive and reliable evaluation, we employed multiple metrics during the training and testing phases. The key novelty of our approach consists in the integration and evaluation of predictive world models within the context of social navigation, as well as in the models themselves. Based on a diverse set of performance metrics, the experimental results provide evidence that predictive world models help improve reinforcement learning techniques for social navigation.

UR - https://link.springer.com/conference/ukci

UR - https://link.springer.com/book/10.1007/978-3-031-47508-5#:~:text=About%20this%20book,UK%20and%20Europe%20held%20annually.

U2 - 10.1007/978-3-031-47508-5_5

DO - 10.1007/978-3-031-47508-5_5

M3 - Conference publication

SN - 978-3-031-47507-8

T3 - Advances in Intelligent Systems and Computing (AISC)

SP - 53

EP - 64

BT - Advances in Computational Intelligence Systems, Contributions Presented at the 22nd UK Workshop on Computational Intelligence

A2 - Naik, Nitin

A2 - Jenkins, Paul

A2 - Grace, Paul

A2 - Yang, Longzhi

A2 - Prajapat, Shaligram

PB - Springer

T2 - The 22nd UK Workship on Computational Intelligence

Y2 - 6 September 2023 through 8 September 2023

ER -

Oguzie G, Ekárt A , Manso LJ. Predictive World Models for Social Navigation. In Naik N, Jenkins P, Grace P, Yang L, Prajapat S, editors, Advances in Computational Intelligence Systems, Contributions Presented at the 22nd UK Workshop on Computational Intelligence. Springer. 2024. p. 53–64. (Advances in Intelligent Systems and Computing (AISC)). doi: 10.1007/978-3-031-47508-5_5

Predictive World Models for Social Navigation

Abstract

Publication series

Conference

Bibliographical note

Access to Document

Other files and links

Embargoed Document

Fingerprint

Cite this