LANGUAGE MODEL CONVERSION DEVICE, SOUND MODEL CONVERSION DEVICE, AND COMPUTER PROGRAM

摘要

PROBLEM TO BE SOLVED: To generate a language model and a sound model which are robust for emphasized utterance of English. SOLUTION: A language model conversion device includes: a syllable border estimation section 92 which estimates syllable borders for respective indexes of a word dictionary; and a phoneme description rewriting section 94 which rewrites a description of phonemes adjacent to each of the syllable borders estimated by the syllable border estimation section to a format wherein a description of ä¾sil} is added to allow insertion of a short-time pause. A sound model is converted to a format similarly allowing a pause to be present at the end of each phoneme model. COPYRIGHT: (C)2006,JPO&NCIPI