Introduction - If you have any usage issues, please Google them yourself
Another common KWS method is Acoustic KWS. Using this method, the engine
does not attempt to transcribe the entire stream of speech. Like the LVCSR-based
method, this method employs the Viterbi search.