Ok, here's how I'd approach the problem. First, load up the dictation engine with the Pronunciation topic, which will return the phonemes spoken by the user ... ... <看更多>