[woodson]

This is about speech to text, not speech recognition.