JANUS: A speech-to-speech translation system using connectionist and symbolic processing strategies

Alex Waibel, Ajay N. Jain, Arthur E. McNair, Hiroaki Saito, Alexander G. Hauptmann, Joe Tebelskis

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

40 Citations (Scopus)

Abstract

The authors present JANUS, a speech-to-speech translation system that utilizes diverse processing strategies including dynamic programming, stochastic techniques, connectionist learning, and traditional AI knowledge representation approaches. JANUS translates continuously spoken English utterances into Japanese and German speech utterances. The overall system performance on a corpus of conference registration conversations is 87%. Two versions of JANUS are compared: one using a LR parser (JANUS1) and one using a connectionist parser (JANUS2). Performance results were mixed, with JANUS1 deriving benefit from a tighter language model and JANUS2 benefitting from greater flexibility.

Original languageEnglish
Title of host publicationProceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing
Editors Anon
PublisherPubl by IEEE
Pages793-796
Number of pages4
Volume2
ISBN (Print)078030033
Publication statusPublished - 1991
Externally publishedYes
EventProceedings of the 1991 International Conference on Acoustics, Speech, and Signal Processing - ICASSP 91 - Toronto, Ont, Can
Duration: 1991 May 141991 May 17

Other

OtherProceedings of the 1991 International Conference on Acoustics, Speech, and Signal Processing - ICASSP 91
CityToronto, Ont, Can
Period91/5/1491/5/17

Fingerprint

Processing
knowledge representation
conversation
dynamic programming
Knowledge representation
Dynamic programming
learning
flexibility

ASJC Scopus subject areas

  • Signal Processing
  • Electrical and Electronic Engineering
  • Acoustics and Ultrasonics

Cite this

Waibel, A., Jain, A. N., McNair, A. E., Saito, H., Hauptmann, A. G., & Tebelskis, J. (1991). JANUS: A speech-to-speech translation system using connectionist and symbolic processing strategies. In Anon (Ed.), Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing (Vol. 2, pp. 793-796). Publ by IEEE.

JANUS : A speech-to-speech translation system using connectionist and symbolic processing strategies. / Waibel, Alex; Jain, Ajay N.; McNair, Arthur E.; Saito, Hiroaki; Hauptmann, Alexander G.; Tebelskis, Joe.

Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing. ed. / Anon. Vol. 2 Publ by IEEE, 1991. p. 793-796.

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

Waibel, A, Jain, AN, McNair, AE, Saito, H, Hauptmann, AG & Tebelskis, J 1991, JANUS: A speech-to-speech translation system using connectionist and symbolic processing strategies. in Anon (ed.), Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing. vol. 2, Publ by IEEE, pp. 793-796, Proceedings of the 1991 International Conference on Acoustics, Speech, and Signal Processing - ICASSP 91, Toronto, Ont, Can, 91/5/14.
Waibel A, Jain AN, McNair AE, Saito H, Hauptmann AG, Tebelskis J. JANUS: A speech-to-speech translation system using connectionist and symbolic processing strategies. In Anon, editor, Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing. Vol. 2. Publ by IEEE. 1991. p. 793-796
Waibel, Alex ; Jain, Ajay N. ; McNair, Arthur E. ; Saito, Hiroaki ; Hauptmann, Alexander G. ; Tebelskis, Joe. / JANUS : A speech-to-speech translation system using connectionist and symbolic processing strategies. Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing. editor / Anon. Vol. 2 Publ by IEEE, 1991. pp. 793-796
@inproceedings{e3a7d1be047544289b6100888717852c,
title = "JANUS: A speech-to-speech translation system using connectionist and symbolic processing strategies",
abstract = "The authors present JANUS, a speech-to-speech translation system that utilizes diverse processing strategies including dynamic programming, stochastic techniques, connectionist learning, and traditional AI knowledge representation approaches. JANUS translates continuously spoken English utterances into Japanese and German speech utterances. The overall system performance on a corpus of conference registration conversations is 87{\%}. Two versions of JANUS are compared: one using a LR parser (JANUS1) and one using a connectionist parser (JANUS2). Performance results were mixed, with JANUS1 deriving benefit from a tighter language model and JANUS2 benefitting from greater flexibility.",
author = "Alex Waibel and Jain, {Ajay N.} and McNair, {Arthur E.} and Hiroaki Saito and Hauptmann, {Alexander G.} and Joe Tebelskis",
year = "1991",
language = "English",
isbn = "078030033",
volume = "2",
pages = "793--796",
editor = "Anon",
booktitle = "Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing",
publisher = "Publ by IEEE",

}

TY - GEN

T1 - JANUS

T2 - A speech-to-speech translation system using connectionist and symbolic processing strategies

AU - Waibel, Alex

AU - Jain, Ajay N.

AU - McNair, Arthur E.

AU - Saito, Hiroaki

AU - Hauptmann, Alexander G.

AU - Tebelskis, Joe

PY - 1991

Y1 - 1991

N2 - The authors present JANUS, a speech-to-speech translation system that utilizes diverse processing strategies including dynamic programming, stochastic techniques, connectionist learning, and traditional AI knowledge representation approaches. JANUS translates continuously spoken English utterances into Japanese and German speech utterances. The overall system performance on a corpus of conference registration conversations is 87%. Two versions of JANUS are compared: one using a LR parser (JANUS1) and one using a connectionist parser (JANUS2). Performance results were mixed, with JANUS1 deriving benefit from a tighter language model and JANUS2 benefitting from greater flexibility.

AB - The authors present JANUS, a speech-to-speech translation system that utilizes diverse processing strategies including dynamic programming, stochastic techniques, connectionist learning, and traditional AI knowledge representation approaches. JANUS translates continuously spoken English utterances into Japanese and German speech utterances. The overall system performance on a corpus of conference registration conversations is 87%. Two versions of JANUS are compared: one using a LR parser (JANUS1) and one using a connectionist parser (JANUS2). Performance results were mixed, with JANUS1 deriving benefit from a tighter language model and JANUS2 benefitting from greater flexibility.

UR - http://www.scopus.com/inward/record.url?scp=0026392360&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0026392360&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:0026392360

SN - 078030033

VL - 2

SP - 793

EP - 796

BT - Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing

A2 - Anon, null

PB - Publ by IEEE

ER -