Improving speech recognition with the Robot Interaction Language

Omar Mubin; Christoph Bartneck; Loe Feijs; Hanneke Hooft van Huysduynen; Jun Hu; Jerry Muelver

doi:10.1089/dst.2012.0010

Improving speech recognition with the Robot Interaction Language

Omar Mubin, Christoph Bartneck, Loe Feijs, Hanneke Hooft van Huysduynen, Jun Hu, Jerry Muelver

Research output: Contribution to journal › Article › peer-review

Abstract

This article presents the design and evaluation of a Robot Interaction Language (ROILA). This speech recognition friendly spoken artificial language is designed to be used by humans for interacting with robots. We evaluated the use of ROILA in a Dutch high school. The language was taught as a part of the science curriculum followed by a controlled experiment, where the language was compared against English. The results from the experiment showed that the ROILA performed better than English on account of both objective recognition accuracy and the subjective assessment by the students. We estimate the trade-off between this benefit in relation to the effort required to learn ROILA. In a regular usage scenario, it would pay off to use ROILA.

Original language	English
Pages (from-to)	79-88
Number of pages	10
Journal	Disruptive Science and Technology
Volume	1
Issue number	2
DOIs	https://doi.org/10.1089/dst.2012.0010
Publication status	Published - 2012

Keywords

speech perception
robots
languages, artificial

Access to Document

10.1089/dst.2012.0010

Cite this

@article{9299739f686c4e56a3e36352fc7f161b,

title = "Improving speech recognition with the Robot Interaction Language",

abstract = "This article presents the design and evaluation of a Robot Interaction Language (ROILA). This speech recognition friendly spoken artificial language is designed to be used by humans for interacting with robots. We evaluated the use of ROILA in a Dutch high school. The language was taught as a part of the science curriculum followed by a controlled experiment, where the language was compared against English. The results from the experiment showed that the ROILA performed better than English on account of both objective recognition accuracy and the subjective assessment by the students. We estimate the trade-off between this benefit in relation to the effort required to learn ROILA. In a regular usage scenario, it would pay off to use ROILA.",

keywords = "speech perception, robots, languages, artificial",

author = "Omar Mubin and Christoph Bartneck and Loe Feijs and {Hooft van Huysduynen}, Hanneke and Jun Hu and Jerry Muelver",

year = "2012",

doi = "10.1089/dst.2012.0010",

language = "English",

volume = "1",

pages = "79--88",

journal = "Disruptive Science and Technology",

issn = "2163-310X",

publisher = "Mary Ann Liebert",

number = "2",

}

TY - JOUR

T1 - Improving speech recognition with the Robot Interaction Language

AU - Mubin, Omar

AU - Bartneck, Christoph

AU - Feijs, Loe

AU - Hooft van Huysduynen, Hanneke

AU - Hu, Jun

AU - Muelver, Jerry

PY - 2012

Y1 - 2012

N2 - This article presents the design and evaluation of a Robot Interaction Language (ROILA). This speech recognition friendly spoken artificial language is designed to be used by humans for interacting with robots. We evaluated the use of ROILA in a Dutch high school. The language was taught as a part of the science curriculum followed by a controlled experiment, where the language was compared against English. The results from the experiment showed that the ROILA performed better than English on account of both objective recognition accuracy and the subjective assessment by the students. We estimate the trade-off between this benefit in relation to the effort required to learn ROILA. In a regular usage scenario, it would pay off to use ROILA.

AB - This article presents the design and evaluation of a Robot Interaction Language (ROILA). This speech recognition friendly spoken artificial language is designed to be used by humans for interacting with robots. We evaluated the use of ROILA in a Dutch high school. The language was taught as a part of the science curriculum followed by a controlled experiment, where the language was compared against English. The results from the experiment showed that the ROILA performed better than English on account of both objective recognition accuracy and the subjective assessment by the students. We estimate the trade-off between this benefit in relation to the effort required to learn ROILA. In a regular usage scenario, it would pay off to use ROILA.

KW - speech perception

KW - robots

KW - languages, artificial

UR - http://handle.uws.edu.au:8081/1959.7/515963

U2 - 10.1089/dst.2012.0010

DO - 10.1089/dst.2012.0010

M3 - Article

SN - 2163-310X

VL - 1

SP - 79

EP - 88

JO - Disruptive Science and Technology

JF - Disruptive Science and Technology

IS - 2

ER -

Improving speech recognition with the Robot Interaction Language

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this