Experience-based imitation using RNNPB

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

Research output: Contribution to journalArticle

17 Citations (Scopus)

Abstract

Robot imitation is a useful and promising alternative to robot programming. Robot imitation involves two crucial issues. The first is how a robot can imitate a human whose physical structure and properties differ greatly from its own. The second is how the robot can generate various motions from finite programmable patterns (generalization). This paper describes a novel approach to robot imitation based on its own physical experiences. We considered the target task of moving an object on a table. For imitation, we focused on an active sensing process in which the robot acquires the relation between the object's motion and its own arm motion. For generalization, we applied the RNNPB (recurrent neural network with parametric bias) model to enable recognition/generation of imitation motions. The robot associates the arm motion which reproduces the observed object's motion presented by a human operator. Experimental results proved the generalization capability of our method, which enables the robot to imitate not only motion it has experienced, but also unknown motion through nonlinear combination of the experienced motions.

Original languageEnglish
Pages (from-to)1351-1367
Number of pages17
JournalAdvanced Robotics
Volume21
Issue number12
DOIs
Publication statusPublished - 2007 Dec 1
Externally publishedYes

Fingerprint

Recurrent neural networks
Robots
Robot programming

Keywords

  • Active sensing
  • Humanoid robot
  • Imitation
  • Recurrent neural network

ASJC Scopus subject areas

  • Control and Systems Engineering

Cite this

Experience-based imitation using RNNPB. / Yokoya, Ryunosuke; Ogata, Tetsuya; Tani, Jun; Komatani, Kazunori; Okuno, Hiroshi G.

In: Advanced Robotics, Vol. 21, No. 12, 01.12.2007, p. 1351-1367.

Research output: Contribution to journalArticle

Yokoya, Ryunosuke ; Ogata, Tetsuya ; Tani, Jun ; Komatani, Kazunori ; Okuno, Hiroshi G. / Experience-based imitation using RNNPB. In: Advanced Robotics. 2007 ; Vol. 21, No. 12. pp. 1351-1367.
@article{bfde5ad3310b4e8d85367d653f81c845,
title = "Experience-based imitation using RNNPB",
abstract = "Robot imitation is a useful and promising alternative to robot programming. Robot imitation involves two crucial issues. The first is how a robot can imitate a human whose physical structure and properties differ greatly from its own. The second is how the robot can generate various motions from finite programmable patterns (generalization). This paper describes a novel approach to robot imitation based on its own physical experiences. We considered the target task of moving an object on a table. For imitation, we focused on an active sensing process in which the robot acquires the relation between the object's motion and its own arm motion. For generalization, we applied the RNNPB (recurrent neural network with parametric bias) model to enable recognition/generation of imitation motions. The robot associates the arm motion which reproduces the observed object's motion presented by a human operator. Experimental results proved the generalization capability of our method, which enables the robot to imitate not only motion it has experienced, but also unknown motion through nonlinear combination of the experienced motions.",
keywords = "Active sensing, Humanoid robot, Imitation, Recurrent neural network",
author = "Ryunosuke Yokoya and Tetsuya Ogata and Jun Tani and Kazunori Komatani and Okuno, {Hiroshi G.}",
year = "2007",
month = "12",
day = "1",
doi = "10.1163/156855307781746106",
language = "English",
volume = "21",
pages = "1351--1367",
journal = "Advanced Robotics",
issn = "0169-1864",
publisher = "Taylor and Francis Ltd.",
number = "12",

}

TY - JOUR

T1 - Experience-based imitation using RNNPB

AU - Yokoya, Ryunosuke

AU - Ogata, Tetsuya

AU - Tani, Jun

AU - Komatani, Kazunori

AU - Okuno, Hiroshi G.

PY - 2007/12/1

Y1 - 2007/12/1

N2 - Robot imitation is a useful and promising alternative to robot programming. Robot imitation involves two crucial issues. The first is how a robot can imitate a human whose physical structure and properties differ greatly from its own. The second is how the robot can generate various motions from finite programmable patterns (generalization). This paper describes a novel approach to robot imitation based on its own physical experiences. We considered the target task of moving an object on a table. For imitation, we focused on an active sensing process in which the robot acquires the relation between the object's motion and its own arm motion. For generalization, we applied the RNNPB (recurrent neural network with parametric bias) model to enable recognition/generation of imitation motions. The robot associates the arm motion which reproduces the observed object's motion presented by a human operator. Experimental results proved the generalization capability of our method, which enables the robot to imitate not only motion it has experienced, but also unknown motion through nonlinear combination of the experienced motions.

AB - Robot imitation is a useful and promising alternative to robot programming. Robot imitation involves two crucial issues. The first is how a robot can imitate a human whose physical structure and properties differ greatly from its own. The second is how the robot can generate various motions from finite programmable patterns (generalization). This paper describes a novel approach to robot imitation based on its own physical experiences. We considered the target task of moving an object on a table. For imitation, we focused on an active sensing process in which the robot acquires the relation between the object's motion and its own arm motion. For generalization, we applied the RNNPB (recurrent neural network with parametric bias) model to enable recognition/generation of imitation motions. The robot associates the arm motion which reproduces the observed object's motion presented by a human operator. Experimental results proved the generalization capability of our method, which enables the robot to imitate not only motion it has experienced, but also unknown motion through nonlinear combination of the experienced motions.

KW - Active sensing

KW - Humanoid robot

KW - Imitation

KW - Recurrent neural network

UR - http://www.scopus.com/inward/record.url?scp=34548449413&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34548449413&partnerID=8YFLogxK

U2 - 10.1163/156855307781746106

DO - 10.1163/156855307781746106

M3 - Article

AN - SCOPUS:34548449413

VL - 21

SP - 1351

EP - 1367

JO - Advanced Robotics

JF - Advanced Robotics

SN - 0169-1864

IS - 12

ER -