JANUS: A Speech-to-speech Translation System Using Connectionist and Symbolic Processing Strategies - Robotics Institute Carnegie Mellon University

JANUS: A Speech-to-speech Translation System Using Connectionist and Symbolic Processing Strategies

Alex Waibel, Ajay N. Jain, Arthur E. McNair, Hideo Saito, Alex Hauptmann, and Joe Tebelskis
Conference Paper, Proceedings of International Conference on Acoustics, Speech, and Signal Processing (ICASSP '91), Vol. 2, pp. 793 - 796, April, 1991

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 (JANUS 1) and one using a connectionist parser (JANUS 2). Performance results were mixed, with JANUS 1 deriving benefit from a tighter language model and JANUS 2 benefitting from greater flexibility.

BibTeX

@conference{Waibel-1991-15817,
author = {Alex Waibel and Ajay N. Jain and Arthur E. McNair and Hideo Saito and Alex Hauptmann and Joe Tebelskis},
title = {JANUS: A Speech-to-speech Translation System Using Connectionist and Symbolic Processing Strategies},
booktitle = {Proceedings of International Conference on Acoustics, Speech, and Signal Processing (ICASSP '91)},
year = {1991},
month = {April},
volume = {2},
pages = {793 - 796},
}