SPEECH RECOGNITION APPARATUS AND RECOGNITION METHOD USING THE SAME

摘要

A voice recognition device and a recognition method using the same are provided to secure a high recognition rate although numerical sounds are continuously inputted by dividing the numerical sound into a pair of semi-syllables and recognizing the divided semi-syllables. One syllable is bisected in a Korean numerical sound by a semi-syllable unit. A rear semi-syllable of a front syllable and a front semi-syllable of a rear syllable are modeled as a pair of semi-syllables in the bisected semi-syllables. The Korean numerical sound inputted from the outside is recognized by using a pair of the semi-syllables. A pair of the semi-syllables include number and unit information.