A Novel Multi-vision Sensor Dataset for Insect-Inspired Outdoor Autonomous Navigation

Jan K.N. Verheyen; Julien Dupeyroux; Guido C.H.E.de Croon

doi:10.1007/978-3-031-20470-8_28

A Novel Multi-vision Sensor Dataset for Insect-Inspired Outdoor Autonomous Navigation

Jan K.N. Verheyen, Julien Dupeyroux, Guido C.H.E.de Croon

Control & Simulation

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

1 Citation (Scopus)

19 Downloads (Pure)

Abstract

Insects have—over millions of years of evolution—perfected many of the systems that roboticists aim to achieve; they can swiftly and robustly navigate through different environments under various conditions while at the same time being highly energy efficient. To reach this level of performance and efficiency, one might want to look at and take inspiration from how these insects achieve their feats. Currently, no dataset exists that allows bio-inspired navigation models to be evaluated over long >100 m real-life routes. We present a novel dataset containing omnidirectional event vision, frame-based vision, depth frames, inertial measurement (IMU) readings, and centimeter-accurate GNSS positioning over kilometer long stretches in and around the TUDelft campus. The dataset is used to evaluate familiarity-based insect-inspired neural navigation models on their performance over longer sequences. It demonstrates that current scene familiarity models are not suited for long-ranged navigation, at least not in their current form.

Original language	English
Title of host publication	Biomimetic and Biohybrid Systems - 11th International Conference, Living Machines 2022, Proceedings
Editors	Alexander Hunt, Vasiliki Vouloutsi, Kenneth Moses, Roger Quinn, Anna Mura, Tony Prescott, Paul F. Verschure
Publisher	Springer
Pages	279-291
Number of pages	13
ISBN (Print)	9783031204692
DOIs	https://doi.org/10.1007/978-3-031-20470-8_28
Publication status	Published - 2022
Event	11th International Conference on Biomimetic and Biohybrid Systems, Living Machines 2022 - Virtual, Online Duration: 19 Jul 2022 → 22 Jul 2022

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	13548 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	11th International Conference on Biomimetic and Biohybrid Systems, Living Machines 2022
City	Virtual, Online
Period	19/07/22 → 22/07/22

Bibliographical note

Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care
Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

Keywords

Event-based camera
GNSS
GPS
Long-range navigation
Neuromorphic systems
RGB Camera

Access to Document

10.1007/978-3-031-20470-8_28

978-3-031-20470-8_28Final published version, 2.16 MB

Cite this

Verheyen, J. K. N., Dupeyroux, J., & Croon, G. C. H. E. D. (2022). A Novel Multi-vision Sensor Dataset for Insect-Inspired Outdoor Autonomous Navigation. In A. Hunt, V. Vouloutsi, K. Moses, R. Quinn, A. Mura, T. Prescott, & P. F. Verschure (Eds.), Biomimetic and Biohybrid Systems - 11th International Conference, Living Machines 2022, Proceedings (pp. 279-291). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13548 LNAI). Springer. https://doi.org/10.1007/978-3-031-20470-8_28

Verheyen, Jan K.N. ; Dupeyroux, Julien ; Croon, Guido C.H.E.de. / A Novel Multi-vision Sensor Dataset for Insect-Inspired Outdoor Autonomous Navigation. Biomimetic and Biohybrid Systems - 11th International Conference, Living Machines 2022, Proceedings. editor / Alexander Hunt ; Vasiliki Vouloutsi ; Kenneth Moses ; Roger Quinn ; Anna Mura ; Tony Prescott ; Paul F. Verschure. Springer, 2022. pp. 279-291 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{f60a4d00d79a4b759bdbe427f3e8328b,

title = "A Novel Multi-vision Sensor Dataset for Insect-Inspired Outdoor Autonomous Navigation",

abstract = "Insects have—over millions of years of evolution—perfected many of the systems that roboticists aim to achieve; they can swiftly and robustly navigate through different environments under various conditions while at the same time being highly energy efficient. To reach this level of performance and efficiency, one might want to look at and take inspiration from how these insects achieve their feats. Currently, no dataset exists that allows bio-inspired navigation models to be evaluated over long >100 m real-life routes. We present a novel dataset containing omnidirectional event vision, frame-based vision, depth frames, inertial measurement (IMU) readings, and centimeter-accurate GNSS positioning over kilometer long stretches in and around the TUDelft campus. The dataset is used to evaluate familiarity-based insect-inspired neural navigation models on their performance over longer sequences. It demonstrates that current scene familiarity models are not suited for long-ranged navigation, at least not in their current form.",

keywords = "Event-based camera, GNSS, GPS, Long-range navigation, Neuromorphic systems, RGB Camera",

author = "Verheyen, {Jan K.N.} and Julien Dupeyroux and Croon, {Guido C.H.E.de}",

note = "Green Open Access added to TU Delft Institutional Repository {\textquoteleft}You share, we take care!{\textquoteright} – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public. ; 11th International Conference on Biomimetic and Biohybrid Systems, Living Machines 2022 ; Conference date: 19-07-2022 Through 22-07-2022",

year = "2022",

doi = "10.1007/978-3-031-20470-8_28",

language = "English",

isbn = "9783031204692",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer",

pages = "279--291",

editor = "Alexander Hunt and Vasiliki Vouloutsi and Kenneth Moses and Roger Quinn and Anna Mura and Tony Prescott and Verschure, {Paul F.}",

booktitle = "Biomimetic and Biohybrid Systems - 11th International Conference, Living Machines 2022, Proceedings",

}

Verheyen, JKN, Dupeyroux, J & Croon, GCHED 2022, A Novel Multi-vision Sensor Dataset for Insect-Inspired Outdoor Autonomous Navigation. in A Hunt, V Vouloutsi, K Moses, R Quinn, A Mura, T Prescott & PF Verschure (eds), Biomimetic and Biohybrid Systems - 11th International Conference, Living Machines 2022, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 13548 LNAI, Springer, pp. 279-291, 11th International Conference on Biomimetic and Biohybrid Systems, Living Machines 2022, Virtual, Online, 19/07/22. https://doi.org/10.1007/978-3-031-20470-8_28

A Novel Multi-vision Sensor Dataset for Insect-Inspired Outdoor Autonomous Navigation. / Verheyen, Jan K.N.; Dupeyroux, Julien; Croon, Guido C.H.E.de.
Biomimetic and Biohybrid Systems - 11th International Conference, Living Machines 2022, Proceedings. ed. / Alexander Hunt; Vasiliki Vouloutsi; Kenneth Moses; Roger Quinn; Anna Mura; Tony Prescott; Paul F. Verschure. Springer, 2022. p. 279-291 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13548 LNAI).

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

TY - GEN

T1 - A Novel Multi-vision Sensor Dataset for Insect-Inspired Outdoor Autonomous Navigation

AU - Verheyen, Jan K.N.

AU - Dupeyroux, Julien

AU - Croon, Guido C.H.E.de

N1 - Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

PY - 2022

Y1 - 2022

N2 - Insects have—over millions of years of evolution—perfected many of the systems that roboticists aim to achieve; they can swiftly and robustly navigate through different environments under various conditions while at the same time being highly energy efficient. To reach this level of performance and efficiency, one might want to look at and take inspiration from how these insects achieve their feats. Currently, no dataset exists that allows bio-inspired navigation models to be evaluated over long >100 m real-life routes. We present a novel dataset containing omnidirectional event vision, frame-based vision, depth frames, inertial measurement (IMU) readings, and centimeter-accurate GNSS positioning over kilometer long stretches in and around the TUDelft campus. The dataset is used to evaluate familiarity-based insect-inspired neural navigation models on their performance over longer sequences. It demonstrates that current scene familiarity models are not suited for long-ranged navigation, at least not in their current form.

AB - Insects have—over millions of years of evolution—perfected many of the systems that roboticists aim to achieve; they can swiftly and robustly navigate through different environments under various conditions while at the same time being highly energy efficient. To reach this level of performance and efficiency, one might want to look at and take inspiration from how these insects achieve their feats. Currently, no dataset exists that allows bio-inspired navigation models to be evaluated over long >100 m real-life routes. We present a novel dataset containing omnidirectional event vision, frame-based vision, depth frames, inertial measurement (IMU) readings, and centimeter-accurate GNSS positioning over kilometer long stretches in and around the TUDelft campus. The dataset is used to evaluate familiarity-based insect-inspired neural navigation models on their performance over longer sequences. It demonstrates that current scene familiarity models are not suited for long-ranged navigation, at least not in their current form.

KW - Event-based camera

KW - GNSS

KW - GPS

KW - Long-range navigation

KW - Neuromorphic systems

KW - RGB Camera

UR - http://www.scopus.com/inward/record.url?scp=85144250976&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-20470-8_28

DO - 10.1007/978-3-031-20470-8_28

M3 - Conference contribution

AN - SCOPUS:85144250976

SN - 9783031204692

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 279

EP - 291

BT - Biomimetic and Biohybrid Systems - 11th International Conference, Living Machines 2022, Proceedings

A2 - Hunt, Alexander

A2 - Vouloutsi, Vasiliki

A2 - Moses, Kenneth

A2 - Quinn, Roger

A2 - Mura, Anna

A2 - Prescott, Tony

A2 - Verschure, Paul F.

PB - Springer

T2 - 11th International Conference on Biomimetic and Biohybrid Systems, Living Machines 2022

Y2 - 19 July 2022 through 22 July 2022

ER -

Verheyen JKN, Dupeyroux J, Croon GCHED. A Novel Multi-vision Sensor Dataset for Insect-Inspired Outdoor Autonomous Navigation. In Hunt A, Vouloutsi V, Moses K, Quinn R, Mura A, Prescott T, Verschure PF, editors, Biomimetic and Biohybrid Systems - 11th International Conference, Living Machines 2022, Proceedings. Springer. 2022. p. 279-291. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-20470-8_28