Speech recognition device, automatic translation device, speech recognition method, program, and data structure

摘要

PPROBLEM TO BE SOLVED: To provide a speech recognition device, for processing speech recognition at high speed. PSOLUTION: The speech recognition device comprises: a sound model storage section 11 in which a context dependent sound model which is learned by setting 23 consonants (b,p,m,f,d,t,n,l,z,c,s,zh,ch,sh,r,j,q,x,g,k,h,ng and a null initial) and 7 vowels (a,a3,i,u,e,o and v) of Chinese language, as a phoneme set; a dictionary information storage section 23 for storing dictionary information including phoneme sequence indicated by the phoneme set which is the same as learning of the sound model, and a Chinese character, in association with each other; a speech recognition section 16 which receives a feature amount extracted from a speech signal generated by utterance, and which acquires sequence of the Chinese character from the feature amount by using at least the sound model and the dictionary information; and an output section 17 for outputting the sequence of the Chinese character as a result of speech recognition. PCOPYRIGHT: (C)2009,JPO&INPIT