Structural feature extraction based on active sensing experiences

Shun Nishide, Tetsuya Ogata, Ryunosuke Yokoya, Kazunori Komatani, Hiroshi G. Okuno, Jun Tani

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

1 Citation (Scopus)

Abstract

Affordance is a feature of an object or environment that implies how to interact with it. Based on affordance theory, humans are said to perceive invariant structures for cognizing the object/environment for generating behaviors. In this paper, the authors present a method to extract invariant structures of objects from visual raw images, based on object manipulation experiences using a humanoid robot. The method consists of two training phases. The first phase utilizes Recurrent Neural Network with Parametric Bias (RNNPB) to self-organize dynamical object features extracted during active sensing with objects. The second phase trains a hierarchical neural network attached to RNNPB for associating object images and robot motions with self-organized object features. Analysis of the model has uncovered static objects features that are closely related to dynamic object motions, such as round or stable.

Original languageEnglish
Title of host publicationProceedings - International Conference on Informatics Education and Research for Knowledge-Circulating Society, ICKS 2008
Pages169-172
Number of pages4
DOIs
Publication statusPublished - 2008
Externally publishedYes
EventInternational Conference on Informatics Education and Research for Knowledge-Circulating Society, ICKS 2008 - Kyoto
Duration: 2008 Jan 172008 Jan 17

Other

OtherInternational Conference on Informatics Education and Research for Knowledge-Circulating Society, ICKS 2008
CityKyoto
Period08/1/1708/1/17

Fingerprint

Recurrent neural networks
Feature extraction
Robots
experience
Neural networks
neural network
robot
trend
manipulation

ASJC Scopus subject areas

  • Information Systems
  • Education

Cite this

Nishide, S., Ogata, T., Yokoya, R., Komatani, K., Okuno, H. G., & Tani, J. (2008). Structural feature extraction based on active sensing experiences. In Proceedings - International Conference on Informatics Education and Research for Knowledge-Circulating Society, ICKS 2008 (pp. 169-172). [4460487] https://doi.org/10.1109/ICKS.2008.9

Structural feature extraction based on active sensing experiences. / Nishide, Shun; Ogata, Tetsuya; Yokoya, Ryunosuke; Komatani, Kazunori; Okuno, Hiroshi G.; Tani, Jun.

Proceedings - International Conference on Informatics Education and Research for Knowledge-Circulating Society, ICKS 2008. 2008. p. 169-172 4460487.

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

Nishide, S, Ogata, T, Yokoya, R, Komatani, K, Okuno, HG & Tani, J 2008, Structural feature extraction based on active sensing experiences. in Proceedings - International Conference on Informatics Education and Research for Knowledge-Circulating Society, ICKS 2008., 4460487, pp. 169-172, International Conference on Informatics Education and Research for Knowledge-Circulating Society, ICKS 2008, Kyoto, 08/1/17. https://doi.org/10.1109/ICKS.2008.9
Nishide S, Ogata T, Yokoya R, Komatani K, Okuno HG, Tani J. Structural feature extraction based on active sensing experiences. In Proceedings - International Conference on Informatics Education and Research for Knowledge-Circulating Society, ICKS 2008. 2008. p. 169-172. 4460487 https://doi.org/10.1109/ICKS.2008.9
Nishide, Shun ; Ogata, Tetsuya ; Yokoya, Ryunosuke ; Komatani, Kazunori ; Okuno, Hiroshi G. ; Tani, Jun. / Structural feature extraction based on active sensing experiences. Proceedings - International Conference on Informatics Education and Research for Knowledge-Circulating Society, ICKS 2008. 2008. pp. 169-172
@inproceedings{c68bd1096e9949d3897806d2bea3d726,
title = "Structural feature extraction based on active sensing experiences",
abstract = "Affordance is a feature of an object or environment that implies how to interact with it. Based on affordance theory, humans are said to perceive invariant structures for cognizing the object/environment for generating behaviors. In this paper, the authors present a method to extract invariant structures of objects from visual raw images, based on object manipulation experiences using a humanoid robot. The method consists of two training phases. The first phase utilizes Recurrent Neural Network with Parametric Bias (RNNPB) to self-organize dynamical object features extracted during active sensing with objects. The second phase trains a hierarchical neural network attached to RNNPB for associating object images and robot motions with self-organized object features. Analysis of the model has uncovered static objects features that are closely related to dynamic object motions, such as round or stable.",
author = "Shun Nishide and Tetsuya Ogata and Ryunosuke Yokoya and Kazunori Komatani and Okuno, {Hiroshi G.} and Jun Tani",
year = "2008",
doi = "10.1109/ICKS.2008.9",
language = "English",
isbn = "0769531288",
pages = "169--172",
booktitle = "Proceedings - International Conference on Informatics Education and Research for Knowledge-Circulating Society, ICKS 2008",

}

TY - GEN

T1 - Structural feature extraction based on active sensing experiences

AU - Nishide, Shun

AU - Ogata, Tetsuya

AU - Yokoya, Ryunosuke

AU - Komatani, Kazunori

AU - Okuno, Hiroshi G.

AU - Tani, Jun

PY - 2008

Y1 - 2008

N2 - Affordance is a feature of an object or environment that implies how to interact with it. Based on affordance theory, humans are said to perceive invariant structures for cognizing the object/environment for generating behaviors. In this paper, the authors present a method to extract invariant structures of objects from visual raw images, based on object manipulation experiences using a humanoid robot. The method consists of two training phases. The first phase utilizes Recurrent Neural Network with Parametric Bias (RNNPB) to self-organize dynamical object features extracted during active sensing with objects. The second phase trains a hierarchical neural network attached to RNNPB for associating object images and robot motions with self-organized object features. Analysis of the model has uncovered static objects features that are closely related to dynamic object motions, such as round or stable.

AB - Affordance is a feature of an object or environment that implies how to interact with it. Based on affordance theory, humans are said to perceive invariant structures for cognizing the object/environment for generating behaviors. In this paper, the authors present a method to extract invariant structures of objects from visual raw images, based on object manipulation experiences using a humanoid robot. The method consists of two training phases. The first phase utilizes Recurrent Neural Network with Parametric Bias (RNNPB) to self-organize dynamical object features extracted during active sensing with objects. The second phase trains a hierarchical neural network attached to RNNPB for associating object images and robot motions with self-organized object features. Analysis of the model has uncovered static objects features that are closely related to dynamic object motions, such as round or stable.

UR - http://www.scopus.com/inward/record.url?scp=50149098358&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=50149098358&partnerID=8YFLogxK

U2 - 10.1109/ICKS.2008.9

DO - 10.1109/ICKS.2008.9

M3 - Conference contribution

AN - SCOPUS:50149098358

SN - 0769531288

SN - 9780769531281

SP - 169

EP - 172

BT - Proceedings - International Conference on Informatics Education and Research for Knowledge-Circulating Society, ICKS 2008

ER -