First European Conference on Speech Communication and Technology

Paris, France
September 27-29, 1989

The Syntax-Oriented Speech Understanding System - SPOJUS-SYNO

Seiichi Nakagawa, Yoshihisa Ohguro, Yasuhide Hashimoto

Toyohashi University of Technology, Toyohashi, Tenpakucho, Japan

This paper describes a syntax oriented spoken Japanese understanding system named "SPOJUS-SYNO". At first, this system makes word HMMs automatically by concatenating syllable-based (trained) HMMs. Then a word lattice is hypothesized by using a word spotting algorithm and word-based HMMs. Finally, the time- synchronous left~to-right parsing algorithm is executed to find the best word sequence from the word lattice according to syntactic knowledge represented by CFG. This system was implemented in the "UNIX-QA" task with the vocabulary size of 521 words. Experimental result shows that the sentence understanding rate was about 80% for six male speakers.

Full Paper

Bibliographic reference.  Nakagawa, Seiichi / Ohguro, Yoshihisa / Hashimoto, Yasuhide (1989): "The syntax-oriented speech understanding system - SPOJUS-SYNO", In EUROSPEECH-1989, 2224-2227.