Analysis of motion searching based on reliable predictability using recurrent neural network

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

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

Abstract

Reliable predictability is one of the main factors that determine human behaviors. The authors developed a model that searches and generates robot motions based on reliable predictability. Training of the model consists of three phases. In the first phase, the model trains a sequential learner, namely Recurrent Neural Network with Parametric Bias, to self-organize robot and object dynamics. In the second phase, Steepest Descent Method is utilized to search for robot motion that induces the most predictable object motion. In the third phase, a hierarchical neural network is trained to link object image with the searched motion. Experiments were conducted with cylindrical objects. Analysis of the results have shown that the robot has acquired the most reliable robot motion, shifting it according to the posture of the object. Twenty motion generation experiments have resulted in generation of robot motion that induces consistent rolling motion of the objects.

Original languageEnglish
Title of host publicationIEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM
Pages192-197
Number of pages6
DOIs
Publication statusPublished - 2009
Externally publishedYes
Event2009 IEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM 2009 - Singapore
Duration: 2009 Jul 142009 Jul 17

Other

Other2009 IEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM 2009
CitySingapore
Period09/7/1409/7/17

Fingerprint

Recurrent neural networks
Robots
Steepest descent method
Experiments
Neural networks

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Control and Systems Engineering
  • Computer Science Applications
  • Software

Cite this

Nishide, S., Ogata, T., Tani, J., Komatani, K., & Okuno, H. G. (2009). Analysis of motion searching based on reliable predictability using recurrent neural network. In IEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM (pp. 192-197). [5230015] https://doi.org/10.1109/AIM.2009.5230015

Analysis of motion searching based on reliable predictability using recurrent neural network. / Nishide, Shun; Ogata, Tetsuya; Tani, Jun; Komatani, Kazunori; Okuno, Hiroshi G.

IEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM. 2009. p. 192-197 5230015.

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

Nishide, S, Ogata, T, Tani, J, Komatani, K & Okuno, HG 2009, Analysis of motion searching based on reliable predictability using recurrent neural network. in IEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM., 5230015, pp. 192-197, 2009 IEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM 2009, Singapore, 09/7/14. https://doi.org/10.1109/AIM.2009.5230015
Nishide S, Ogata T, Tani J, Komatani K, Okuno HG. Analysis of motion searching based on reliable predictability using recurrent neural network. In IEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM. 2009. p. 192-197. 5230015 https://doi.org/10.1109/AIM.2009.5230015
Nishide, Shun ; Ogata, Tetsuya ; Tani, Jun ; Komatani, Kazunori ; Okuno, Hiroshi G. / Analysis of motion searching based on reliable predictability using recurrent neural network. IEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM. 2009. pp. 192-197
@inproceedings{6ed05dd2d69e4d4f8a14232b47826fea,
title = "Analysis of motion searching based on reliable predictability using recurrent neural network",
abstract = "Reliable predictability is one of the main factors that determine human behaviors. The authors developed a model that searches and generates robot motions based on reliable predictability. Training of the model consists of three phases. In the first phase, the model trains a sequential learner, namely Recurrent Neural Network with Parametric Bias, to self-organize robot and object dynamics. In the second phase, Steepest Descent Method is utilized to search for robot motion that induces the most predictable object motion. In the third phase, a hierarchical neural network is trained to link object image with the searched motion. Experiments were conducted with cylindrical objects. Analysis of the results have shown that the robot has acquired the most reliable robot motion, shifting it according to the posture of the object. Twenty motion generation experiments have resulted in generation of robot motion that induces consistent rolling motion of the objects.",
author = "Shun Nishide and Tetsuya Ogata and Jun Tani and Kazunori Komatani and Okuno, {Hiroshi G.}",
year = "2009",
doi = "10.1109/AIM.2009.5230015",
language = "English",
isbn = "9781424428533",
pages = "192--197",
booktitle = "IEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM",

}

TY - GEN

T1 - Analysis of motion searching based on reliable predictability using recurrent neural network

AU - Nishide, Shun

AU - Ogata, Tetsuya

AU - Tani, Jun

AU - Komatani, Kazunori

AU - Okuno, Hiroshi G.

PY - 2009

Y1 - 2009

N2 - Reliable predictability is one of the main factors that determine human behaviors. The authors developed a model that searches and generates robot motions based on reliable predictability. Training of the model consists of three phases. In the first phase, the model trains a sequential learner, namely Recurrent Neural Network with Parametric Bias, to self-organize robot and object dynamics. In the second phase, Steepest Descent Method is utilized to search for robot motion that induces the most predictable object motion. In the third phase, a hierarchical neural network is trained to link object image with the searched motion. Experiments were conducted with cylindrical objects. Analysis of the results have shown that the robot has acquired the most reliable robot motion, shifting it according to the posture of the object. Twenty motion generation experiments have resulted in generation of robot motion that induces consistent rolling motion of the objects.

AB - Reliable predictability is one of the main factors that determine human behaviors. The authors developed a model that searches and generates robot motions based on reliable predictability. Training of the model consists of three phases. In the first phase, the model trains a sequential learner, namely Recurrent Neural Network with Parametric Bias, to self-organize robot and object dynamics. In the second phase, Steepest Descent Method is utilized to search for robot motion that induces the most predictable object motion. In the third phase, a hierarchical neural network is trained to link object image with the searched motion. Experiments were conducted with cylindrical objects. Analysis of the results have shown that the robot has acquired the most reliable robot motion, shifting it according to the posture of the object. Twenty motion generation experiments have resulted in generation of robot motion that induces consistent rolling motion of the objects.

UR - http://www.scopus.com/inward/record.url?scp=70350462086&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70350462086&partnerID=8YFLogxK

U2 - 10.1109/AIM.2009.5230015

DO - 10.1109/AIM.2009.5230015

M3 - Conference contribution

SN - 9781424428533

SP - 192

EP - 197

BT - IEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM

ER -