Planar Pose Estimation Using Object Detection and Reinforcement Learning

Frederik Nørby Rasmussen; Sebastian Terp Andersen; Bjarne Grossmann; Evangelos Boukas; Lazaros Nalpantidis

doi:10.1007/978-3-030-34995-0_32

Planar Pose Estimation Using Object Detection and Reinforcement Learning

Frederik Nørby Rasmussen, Sebastian Terp Andersen, Bjarne Grossmann, Evangelos Boukas, Lazaros Nalpantidis^*

^*Corresponding author for this work

Aalborg University

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

Abstract

Pose estimation concerns systems or models dealing with the determination of a static object’s pose using, in this case, vision. This paper approaching the problem with an active vision-based solution, that integrates both perception and action in the same model. The problem is solved using a combination of neural networks for object detection and a reinforcement learning architecture for moving a camera and estimating the pose. A robotic implementation of the proposed active vision system is used for testing with promising results. Experiments show that our approach does not only solve the simple task of planar visual pose estimation, but also exhibits robustness to changes in the environment.

Original language	English
Title of host publication	Proceedings of 12th International Conference on Computer Vision Systems
Editors	Dimitrios Tzovaras, Dimitrios Giakoumis, Markus Vincze, Antonis Argyros
Publisher	Springer
Publication date	1 Jan 2019
Pages	353-365
ISBN (Print)	9783030349943
DOIs	https://doi.org/10.1007/978-3-030-34995-0_32
Publication status	Published - 1 Jan 2019
Event	12th International Conference on Computer Vision Systems - Thessaloniki, Greece Duration: 23 Sep 2019 → 25 Sep 2019 Conference number: 12

Conference

Conference	12th International Conference on Computer Vision Systems
Number	12
Country/Territory	Greece
City	Thessaloniki
Period	23/09/2019 → 25/09/2019

Series	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11754 LNCS
ISSN	0302-9743

Keywords

Object detection
Pose estimation
Reinforcement learning

Access to Document

10.1007/978-3-030-34995-0_32

OpenUrl availability

Full text

Cite this

Nørby Rasmussen, F., Terp Andersen, S., Grossmann, B., Boukas, E., & Nalpantidis, L. (2019). Planar Pose Estimation Using Object Detection and Reinforcement Learning. In D. Tzovaras, D. Giakoumis, M. Vincze, & A. Argyros (Eds.), Proceedings of 12th International Conference on Computer Vision Systems (pp. 353-365). Springer. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) Vol. 11754 LNCS https://doi.org/10.1007/978-3-030-34995-0_32

Nørby Rasmussen, Frederik ; Terp Andersen, Sebastian ; Grossmann, Bjarne et al. / Planar Pose Estimation Using Object Detection and Reinforcement Learning. Proceedings of 12th International Conference on Computer Vision Systems. editor / Dimitrios Tzovaras ; Dimitrios Giakoumis ; Markus Vincze ; Antonis Argyros. Springer, 2019. pp. 353-365 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 11754 LNCS).

@inproceedings{feca2e4172f44b679cbf449f5e807846,

title = "Planar Pose Estimation Using Object Detection and Reinforcement Learning",

abstract = "Pose estimation concerns systems or models dealing with the determination of a static object{\textquoteright}s pose using, in this case, vision. This paper approaching the problem with an active vision-based solution, that integrates both perception and action in the same model. The problem is solved using a combination of neural networks for object detection and a reinforcement learning architecture for moving a camera and estimating the pose. A robotic implementation of the proposed active vision system is used for testing with promising results. Experiments show that our approach does not only solve the simple task of planar visual pose estimation, but also exhibits robustness to changes in the environment.",

keywords = "Object detection, Pose estimation, Reinforcement learning",

author = "{N{\o}rby Rasmussen}, Frederik and {Terp Andersen}, Sebastian and Bjarne Grossmann and Evangelos Boukas and Lazaros Nalpantidis",

year = "2019",

month = jan,

day = "1",

doi = "10.1007/978-3-030-34995-0_32",

language = "English",

isbn = "9783030349943",

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

publisher = "Springer",

pages = "353--365",

editor = "Dimitrios Tzovaras and Dimitrios Giakoumis and Markus Vincze and Antonis Argyros",

booktitle = "Proceedings of 12th International Conference on Computer Vision Systems",

note = "12th International Conference on Computer Vision Systems, ICVS 2019 ; Conference date: 23-09-2019 Through 25-09-2019",

}

Nørby Rasmussen, F, Terp Andersen, S, Grossmann, B, Boukas, E & Nalpantidis, L 2019, Planar Pose Estimation Using Object Detection and Reinforcement Learning. in D Tzovaras, D Giakoumis, M Vincze & A Argyros (eds), Proceedings of 12th International Conference on Computer Vision Systems. Springer, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11754 LNCS, pp. 353-365, 12th International Conference on Computer Vision Systems, Thessaloniki, Greece, 23/09/2019. https://doi.org/10.1007/978-3-030-34995-0_32

Planar Pose Estimation Using Object Detection and Reinforcement Learning. / Nørby Rasmussen, Frederik; Terp Andersen, Sebastian; Grossmann, Bjarne et al.

Proceedings of 12th International Conference on Computer Vision Systems. ed. / Dimitrios Tzovaras; Dimitrios Giakoumis; Markus Vincze; Antonis Argyros. Springer, 2019. p. 353-365 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 11754 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

TY - GEN

T1 - Planar Pose Estimation Using Object Detection and Reinforcement Learning

AU - Nørby Rasmussen, Frederik

AU - Terp Andersen, Sebastian

AU - Grossmann, Bjarne

AU - Boukas, Evangelos

AU - Nalpantidis, Lazaros

N1 - Conference code: 12

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Pose estimation concerns systems or models dealing with the determination of a static object’s pose using, in this case, vision. This paper approaching the problem with an active vision-based solution, that integrates both perception and action in the same model. The problem is solved using a combination of neural networks for object detection and a reinforcement learning architecture for moving a camera and estimating the pose. A robotic implementation of the proposed active vision system is used for testing with promising results. Experiments show that our approach does not only solve the simple task of planar visual pose estimation, but also exhibits robustness to changes in the environment.

AB - Pose estimation concerns systems or models dealing with the determination of a static object’s pose using, in this case, vision. This paper approaching the problem with an active vision-based solution, that integrates both perception and action in the same model. The problem is solved using a combination of neural networks for object detection and a reinforcement learning architecture for moving a camera and estimating the pose. A robotic implementation of the proposed active vision system is used for testing with promising results. Experiments show that our approach does not only solve the simple task of planar visual pose estimation, but also exhibits robustness to changes in the environment.

KW - Object detection

KW - Pose estimation

KW - Reinforcement learning

U2 - 10.1007/978-3-030-34995-0_32

DO - 10.1007/978-3-030-34995-0_32

M3 - Article in proceedings

AN - SCOPUS:85076986438

SN - 9783030349943

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

SP - 353

EP - 365

BT - Proceedings of 12th International Conference on Computer Vision Systems

A2 - Tzovaras, Dimitrios

A2 - Giakoumis, Dimitrios

A2 - Vincze, Markus

A2 - Argyros, Antonis

PB - Springer

T2 - 12th International Conference on Computer Vision Systems

Y2 - 23 September 2019 through 25 September 2019

ER -

Nørby Rasmussen F, Terp Andersen S, Grossmann B, Boukas E , Nalpantidis L. Planar Pose Estimation Using Object Detection and Reinforcement Learning. In Tzovaras D, Giakoumis D, Vincze M, Argyros A, editors, Proceedings of 12th International Conference on Computer Vision Systems. Springer. 2019. p. 353-365. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 11754 LNCS). doi: 10.1007/978-3-030-34995-0_32

Planar Pose Estimation Using Object Detection and Reinforcement Learning

Abstract

Conference

Keywords

Access to Document

OpenUrl availability

Fingerprint

Cite this