speech_recognition
speech_recognition copied to clipboard
Moved energy_threshold into an optional Recognizer() __init__ param.
Explanation: Between the hard-coded default of 300, and the automatic adjustment, is the ability to let the caller pick the value. In our case, our environment has people talking, and while fiddling with the damping and such, for the asymmetric moving avg, also should be moved into configurable params, I figured this was at least a start (and helped in our situation). :)