Tampere University of Technology

TUTCRIS Research Portal

Optimizing gaze direction in a visual navigation task

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Standard

Optimizing gaze direction in a visual navigation task. / Välimäki, Tuomas; Ritala, Risto.

2016 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, 2016. p. 1427-1432.

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Harvard

Välimäki, T & Ritala, R 2016, Optimizing gaze direction in a visual navigation task. in 2016 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, pp. 1427-1432, IEEE International Conference on Robotics and Automation, 1/01/00. https://doi.org/10.1109/ICRA.2016.7487276

APA

Välimäki, T., & Ritala, R. (2016). Optimizing gaze direction in a visual navigation task. In 2016 IEEE International Conference on Robotics and Automation (ICRA) (pp. 1427-1432). IEEE. https://doi.org/10.1109/ICRA.2016.7487276

Vancouver

Välimäki T, Ritala R. Optimizing gaze direction in a visual navigation task. In 2016 IEEE International Conference on Robotics and Automation (ICRA) . IEEE. 2016. p. 1427-1432 https://doi.org/10.1109/ICRA.2016.7487276

Author

Välimäki, Tuomas ; Ritala, Risto. / Optimizing gaze direction in a visual navigation task. 2016 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, 2016. pp. 1427-1432

Bibtex - Download

@inproceedings{132046c899df479fa0f317aa063a8439,
title = "Optimizing gaze direction in a visual navigation task",
abstract = "Navigation in an unknown environment consists of multiple separable subtasks, such as collecting information about the surroundings and navigating to the current goal. In the case of pure visual navigation, all these subtasks need to utilize the same vision system, and therefore a way to optimally control the direction of focus is needed. We present a case study, where we model the active sensing problem of directing the gaze of a mobile robot with three machine vision cameras as a partially observable Markov decision process (POMDP) using a mutual information (MI) based reward function. The key aspect of the solution is that the cameras are dynamically used either in monocular or stereo configuration. The benefits of using the proposed active sensing implementation are demonstrated with simulations and experiments on a real robot.",
author = "Tuomas V{\"a}lim{\"a}ki and Risto Ritala",
year = "2016",
month = "6",
day = "8",
doi = "10.1109/ICRA.2016.7487276",
language = "English",
isbn = "9781467380263",
publisher = "IEEE",
pages = "1427--1432",
booktitle = "2016 IEEE International Conference on Robotics and Automation (ICRA)",

}

RIS (suitable for import to EndNote) - Download

TY - GEN

T1 - Optimizing gaze direction in a visual navigation task

AU - Välimäki, Tuomas

AU - Ritala, Risto

PY - 2016/6/8

Y1 - 2016/6/8

N2 - Navigation in an unknown environment consists of multiple separable subtasks, such as collecting information about the surroundings and navigating to the current goal. In the case of pure visual navigation, all these subtasks need to utilize the same vision system, and therefore a way to optimally control the direction of focus is needed. We present a case study, where we model the active sensing problem of directing the gaze of a mobile robot with three machine vision cameras as a partially observable Markov decision process (POMDP) using a mutual information (MI) based reward function. The key aspect of the solution is that the cameras are dynamically used either in monocular or stereo configuration. The benefits of using the proposed active sensing implementation are demonstrated with simulations and experiments on a real robot.

AB - Navigation in an unknown environment consists of multiple separable subtasks, such as collecting information about the surroundings and navigating to the current goal. In the case of pure visual navigation, all these subtasks need to utilize the same vision system, and therefore a way to optimally control the direction of focus is needed. We present a case study, where we model the active sensing problem of directing the gaze of a mobile robot with three machine vision cameras as a partially observable Markov decision process (POMDP) using a mutual information (MI) based reward function. The key aspect of the solution is that the cameras are dynamically used either in monocular or stereo configuration. The benefits of using the proposed active sensing implementation are demonstrated with simulations and experiments on a real robot.

U2 - 10.1109/ICRA.2016.7487276

DO - 10.1109/ICRA.2016.7487276

M3 - Conference contribution

SN - 9781467380263

SP - 1427

EP - 1432

BT - 2016 IEEE International Conference on Robotics and Automation (ICRA)

PB - IEEE

ER -