currently vad detection happens only when gather verb's minBargeinWordCount is 0, we have vad at config verb, we should start vad detection when config.vad.enable is true
The purpose of this is to start streaming audio to asr when vad is detected specially with Microsoft, which send garbage Yes or No if it get empty audio stream
currently vad detection happens only when gather verb's minBargeinWordCount is 0, we have vad at config verb, we should start vad detection when config.vad.enable is true