SPEECH RECOGNITION METHOD AND APPARATUS, SPEECH RECOGNITION ENGINE GENERATION METHOD AND APPARATUS BASED ON PRONOUNCE SIMILARITY

摘要

Audio recognition method is in view of pronunciation similarity and relevant apparatus. This method comprises the following steps: receiving the user of voice input; Aligned phoneme sequence is sent, is inputted from the voice, speech recognition engine is arrived; Speech recognition is received by speech recognition engine. The speech recognition engine provides speech recognition result distance based on pronunciation phonemes sequence, inputs from the voice. ;The 2016 of copyright KIPO submissions