Graph Neural Networks for Human-aware Social Navigation

Luis J. Manso; Ronit R. Jorvekar; Diego Faria; Pablo Bustos; Pilar Bachiller

doi:10.1007/978-3-030-62579-5_12

Graph Neural Networks for Human-aware Social Navigation

Luis J. Manso^*, Ronit R. Jorvekar, Diego Faria, Pablo Bustos, Pilar Bachiller

^*Corresponding author for this work

College of Engineering and Physical Sciences

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

Abstract

Autonomous navigation is a key skill for assistive and service robots. To be successful, robots have to comply with social rules, such as avoiding the personal spaces of the people surrounding them, or not getting in the way of human-to-human and human-to-object interactions. This paper suggests using Graph Neural Networks to model how inconvenient the presence of a robot would be in a particular scenario according to learned human conventions so that it can be used by path planning algorithms. To do so, we propose two automated scenario-to-graph transformations and benchmark them with different Graph Neural Networks using the SocNav1 dataset. We achieve close-to-human performance in the dataset and argue that, in addition to its promising results, the main advantage of the approach is its scalability in terms of the number of social factors that can be considered and easily embedded in code in comparison with model-based approaches. The code used to train and test the resulting graph neural network is available in a public repository.

Original language	English
Title of host publication	Advances in Physical Agents II - Proceedings of the 21st International Workshop of Physical Agents WAF 2020
Editors	Luis M. Bergasa, Manuel Ocaña, Rafael Barea, Elena López-Guillén, Pedro Revenga
Publisher	Springer
Pages	167-179
Number of pages	13
ISBN (Electronic)	978-3-030-62579-5
ISBN (Print)	978-3-030-62578-8
DOIs	https://doi.org/10.1007/978-3-030-62579-5_12
Publication status	Published - 3 Nov 2020
Event	21st International Workshop of Physical Agents (WAF2020) - Madrid, Spain Duration: 19 Nov 2020 → 20 Nov 2020

Publication series

Name	Advances in Intelligent Systems and Computing
Volume	1285
ISSN (Print)	2194-5357
ISSN (Electronic)	2194-5365

Conference

Conference	21st International Workshop of Physical Agents (WAF2020)
Country/Territory	Spain
City	Madrid
Period	19/11/20 → 20/11/20

Bibliographical note

Keywords

Graph Neural Networks
Human-robot interaction
Social navigation

Access to Document

10.1007/978-3-030-62579-5_12

Graph Neural Networks for Human-aware Social Navigation
© 2020 The Authors
Submitted manuscript, 425 KB

https://arxiv.org/abs/1909.09003Licence: CC BY-NC-SA 3.0

Cite this

Manso, L. J., Jorvekar, R. R., Faria, D., Bustos, P., & Bachiller, P. (2020). Graph Neural Networks for Human-aware Social Navigation. In L. M. Bergasa, M. Ocaña, R. Barea, E. López-Guillén, & P. Revenga (Eds.), Advances in Physical Agents II - Proceedings of the 21st International Workshop of Physical Agents WAF 2020 (pp. 167-179). (Advances in Intelligent Systems and Computing; Vol. 1285). Springer. https://doi.org/10.1007/978-3-030-62579-5_12

Manso, Luis J. ; Jorvekar, Ronit R. ; Faria, Diego et al. / Graph Neural Networks for Human-aware Social Navigation. Advances in Physical Agents II - Proceedings of the 21st International Workshop of Physical Agents WAF 2020. editor / Luis M. Bergasa ; Manuel Ocaña ; Rafael Barea ; Elena López-Guillén ; Pedro Revenga. Springer, 2020. pp. 167-179 (Advances in Intelligent Systems and Computing).

@inproceedings{259fd0e33c8c4201aa2f5a9115a54fc0,

title = "Graph Neural Networks for Human-aware Social Navigation",

abstract = "Autonomous navigation is a key skill for assistive and service robots. To be successful, robots have to comply with social rules, such as avoiding the personal spaces of the people surrounding them, or not getting in the way of human-to-human and human-to-object interactions. This paper suggests using Graph Neural Networks to model how inconvenient the presence of a robot would be in a particular scenario according to learned human conventions so that it can be used by path planning algorithms. To do so, we propose two automated scenario-to-graph transformations and benchmark them with different Graph Neural Networks using the SocNav1 dataset. We achieve close-to-human performance in the dataset and argue that, in addition to its promising results, the main advantage of the approach is its scalability in terms of the number of social factors that can be considered and easily embedded in code in comparison with model-based approaches. The code used to train and test the resulting graph neural network is available in a public repository.",

keywords = "Graph Neural Networks, Human-robot interaction, Social navigation",

author = "Manso, {Luis J.} and Jorvekar, {Ronit R.} and Diego Faria and Pablo Bustos and Pilar Bachiller",

note = "{\textcopyright} 2020 The Authors; 21st International Workshop of Physical Agents (WAF2020) ; Conference date: 19-11-2020 Through 20-11-2020",

year = "2020",

month = nov,

day = "3",

doi = "10.1007/978-3-030-62579-5_12",

language = "English",

isbn = "978-3-030-62578-8",

series = "Advances in Intelligent Systems and Computing",

publisher = "Springer",

pages = "167--179",

editor = "Bergasa, {Luis M.} and Manuel Oca{\~n}a and Rafael Barea and Elena L{\'o}pez-Guill{\'e}n and Pedro Revenga",

booktitle = "Advances in Physical Agents II - Proceedings of the 21st International Workshop of Physical Agents WAF 2020",

address = "Germany",

}

Manso, LJ, Jorvekar, RR, Faria, D, Bustos, P & Bachiller, P 2020, Graph Neural Networks for Human-aware Social Navigation. in LM Bergasa, M Ocaña, R Barea, E López-Guillén & P Revenga (eds), Advances in Physical Agents II - Proceedings of the 21st International Workshop of Physical Agents WAF 2020. Advances in Intelligent Systems and Computing, vol. 1285, Springer, pp. 167-179, 21st International Workshop of Physical Agents (WAF2020), Madrid, Spain, 19/11/20. https://doi.org/10.1007/978-3-030-62579-5_12

Graph Neural Networks for Human-aware Social Navigation. / Manso, Luis J.; Jorvekar, Ronit R.; Faria, Diego et al.
Advances in Physical Agents II - Proceedings of the 21st International Workshop of Physical Agents WAF 2020. ed. / Luis M. Bergasa; Manuel Ocaña; Rafael Barea; Elena López-Guillén; Pedro Revenga. Springer, 2020. p. 167-179 (Advances in Intelligent Systems and Computing; Vol. 1285).

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

TY - GEN

T1 - Graph Neural Networks for Human-aware Social Navigation

AU - Manso, Luis J.

AU - Jorvekar, Ronit R.

AU - Faria, Diego

AU - Bustos, Pablo

AU - Bachiller, Pilar

PY - 2020/11/3

Y1 - 2020/11/3

N2 - Autonomous navigation is a key skill for assistive and service robots. To be successful, robots have to comply with social rules, such as avoiding the personal spaces of the people surrounding them, or not getting in the way of human-to-human and human-to-object interactions. This paper suggests using Graph Neural Networks to model how inconvenient the presence of a robot would be in a particular scenario according to learned human conventions so that it can be used by path planning algorithms. To do so, we propose two automated scenario-to-graph transformations and benchmark them with different Graph Neural Networks using the SocNav1 dataset. We achieve close-to-human performance in the dataset and argue that, in addition to its promising results, the main advantage of the approach is its scalability in terms of the number of social factors that can be considered and easily embedded in code in comparison with model-based approaches. The code used to train and test the resulting graph neural network is available in a public repository.

AB - Autonomous navigation is a key skill for assistive and service robots. To be successful, robots have to comply with social rules, such as avoiding the personal spaces of the people surrounding them, or not getting in the way of human-to-human and human-to-object interactions. This paper suggests using Graph Neural Networks to model how inconvenient the presence of a robot would be in a particular scenario according to learned human conventions so that it can be used by path planning algorithms. To do so, we propose two automated scenario-to-graph transformations and benchmark them with different Graph Neural Networks using the SocNav1 dataset. We achieve close-to-human performance in the dataset and argue that, in addition to its promising results, the main advantage of the approach is its scalability in terms of the number of social factors that can be considered and easily embedded in code in comparison with model-based approaches. The code used to train and test the resulting graph neural network is available in a public repository.

KW - Graph Neural Networks

KW - Human-robot interaction

KW - Social navigation

UR - https://link.springer.com/chapter/10.1007/978-3-030-62579-5_12

UR - https://github.com/robocomp/sngnn

UR - http://www.scopus.com/inward/record.url?scp=85097381428&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-62579-5_12

DO - 10.1007/978-3-030-62579-5_12

M3 - Conference publication

SN - 978-3-030-62578-8

T3 - Advances in Intelligent Systems and Computing

SP - 167

EP - 179

BT - Advances in Physical Agents II - Proceedings of the 21st International Workshop of Physical Agents WAF 2020

A2 - Bergasa, Luis M.

A2 - Ocaña, Manuel

A2 - Barea, Rafael

A2 - López-Guillén, Elena

A2 - Revenga, Pedro

PB - Springer

T2 - 21st International Workshop of Physical Agents (WAF2020)

Y2 - 19 November 2020 through 20 November 2020

ER -

Manso LJ, Jorvekar RR, Faria D, Bustos P, Bachiller P. Graph Neural Networks for Human-aware Social Navigation. In Bergasa LM, Ocaña M, Barea R, López-Guillén E, Revenga P, editors, Advances in Physical Agents II - Proceedings of the 21st International Workshop of Physical Agents WAF 2020. Springer. 2020. p. 167-179. (Advances in Intelligent Systems and Computing). doi: 10.1007/978-3-030-62579-5_12

Graph Neural Networks for Human-aware Social Navigation

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Best Paper Award

Cite this

Graph Neural Networks for Human-aware Social Navigation

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Prizes

Best Paper Award

Cite this