Active object recognition on a humanoid robot

Björn Browatzki, Vadim Tikhanoff, Giorgio Metta, Heinrich Bulthoff, Christian Wallraven

Research output: Chapter in Book/Report/Conference proceedingConference contribution

35 Citations (Scopus)

Abstract

Interaction with its environment is a key requisite for a humanoid robot. Especially the ability to recognize and manipulate unknown objects is crucial to successfully work in natural environments. Visual object recognition, however, still remains a challenging problem, as three-dimensional objects often give rise to ambiguous, two-dimensional views. Here, we propose a perception-driven, multisensory exploration and recognition scheme to actively resolve ambiguities that emerge at certain viewpoints. We define an efficient method to acquire two-dimensional views in an object-centered task space and sample characteristic views on a view sphere. Information is accumulated during the recognition process and used to select actions expected to be most beneficial in discriminating similar objects. Besides visual information we take into account proprioceptive information to create more reliable hypotheses. Simulation and real-world results clearly demonstrate the efficiency of active, multisensory exploration over passive, visiononly recognition methods.

Original languageEnglish
Title of host publicationProceedings - IEEE International Conference on Robotics and Automation
Pages2021-2028
Number of pages8
DOIs
Publication statusPublished - 2012 Aug 6

Fingerprint

Object recognition
Robots

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence
  • Control and Systems Engineering
  • Electrical and Electronic Engineering

Cite this

Browatzki, B., Tikhanoff, V., Metta, G., Bulthoff, H., & Wallraven, C. (2012). Active object recognition on a humanoid robot. In Proceedings - IEEE International Conference on Robotics and Automation (pp. 2021-2028). [6225218] https://doi.org/10.1109/ICRA.2012.6225218

Active object recognition on a humanoid robot. / Browatzki, Björn; Tikhanoff, Vadim; Metta, Giorgio; Bulthoff, Heinrich; Wallraven, Christian.

Proceedings - IEEE International Conference on Robotics and Automation. 2012. p. 2021-2028 6225218.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Browatzki, B, Tikhanoff, V, Metta, G, Bulthoff, H & Wallraven, C 2012, Active object recognition on a humanoid robot. in Proceedings - IEEE International Conference on Robotics and Automation., 6225218, pp. 2021-2028. https://doi.org/10.1109/ICRA.2012.6225218
Browatzki B, Tikhanoff V, Metta G, Bulthoff H, Wallraven C. Active object recognition on a humanoid robot. In Proceedings - IEEE International Conference on Robotics and Automation. 2012. p. 2021-2028. 6225218 https://doi.org/10.1109/ICRA.2012.6225218
Browatzki, Björn ; Tikhanoff, Vadim ; Metta, Giorgio ; Bulthoff, Heinrich ; Wallraven, Christian. / Active object recognition on a humanoid robot. Proceedings - IEEE International Conference on Robotics and Automation. 2012. pp. 2021-2028
@inproceedings{a3e14f9440ab4677896737ab50b92d92,
title = "Active object recognition on a humanoid robot",
abstract = "Interaction with its environment is a key requisite for a humanoid robot. Especially the ability to recognize and manipulate unknown objects is crucial to successfully work in natural environments. Visual object recognition, however, still remains a challenging problem, as three-dimensional objects often give rise to ambiguous, two-dimensional views. Here, we propose a perception-driven, multisensory exploration and recognition scheme to actively resolve ambiguities that emerge at certain viewpoints. We define an efficient method to acquire two-dimensional views in an object-centered task space and sample characteristic views on a view sphere. Information is accumulated during the recognition process and used to select actions expected to be most beneficial in discriminating similar objects. Besides visual information we take into account proprioceptive information to create more reliable hypotheses. Simulation and real-world results clearly demonstrate the efficiency of active, multisensory exploration over passive, visiononly recognition methods.",
author = "Bj{\"o}rn Browatzki and Vadim Tikhanoff and Giorgio Metta and Heinrich Bulthoff and Christian Wallraven",
year = "2012",
month = "8",
day = "6",
doi = "10.1109/ICRA.2012.6225218",
language = "English",
isbn = "9781467314039",
pages = "2021--2028",
booktitle = "Proceedings - IEEE International Conference on Robotics and Automation",

}

TY - GEN

T1 - Active object recognition on a humanoid robot

AU - Browatzki, Björn

AU - Tikhanoff, Vadim

AU - Metta, Giorgio

AU - Bulthoff, Heinrich

AU - Wallraven, Christian

PY - 2012/8/6

Y1 - 2012/8/6

N2 - Interaction with its environment is a key requisite for a humanoid robot. Especially the ability to recognize and manipulate unknown objects is crucial to successfully work in natural environments. Visual object recognition, however, still remains a challenging problem, as three-dimensional objects often give rise to ambiguous, two-dimensional views. Here, we propose a perception-driven, multisensory exploration and recognition scheme to actively resolve ambiguities that emerge at certain viewpoints. We define an efficient method to acquire two-dimensional views in an object-centered task space and sample characteristic views on a view sphere. Information is accumulated during the recognition process and used to select actions expected to be most beneficial in discriminating similar objects. Besides visual information we take into account proprioceptive information to create more reliable hypotheses. Simulation and real-world results clearly demonstrate the efficiency of active, multisensory exploration over passive, visiononly recognition methods.

AB - Interaction with its environment is a key requisite for a humanoid robot. Especially the ability to recognize and manipulate unknown objects is crucial to successfully work in natural environments. Visual object recognition, however, still remains a challenging problem, as three-dimensional objects often give rise to ambiguous, two-dimensional views. Here, we propose a perception-driven, multisensory exploration and recognition scheme to actively resolve ambiguities that emerge at certain viewpoints. We define an efficient method to acquire two-dimensional views in an object-centered task space and sample characteristic views on a view sphere. Information is accumulated during the recognition process and used to select actions expected to be most beneficial in discriminating similar objects. Besides visual information we take into account proprioceptive information to create more reliable hypotheses. Simulation and real-world results clearly demonstrate the efficiency of active, multisensory exploration over passive, visiononly recognition methods.

UR - http://www.scopus.com/inward/record.url?scp=84864487876&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84864487876&partnerID=8YFLogxK

U2 - 10.1109/ICRA.2012.6225218

DO - 10.1109/ICRA.2012.6225218

M3 - Conference contribution

SN - 9781467314039

SP - 2021

EP - 2028

BT - Proceedings - IEEE International Conference on Robotics and Automation

ER -