Open ankitj05 opened 6 years ago
you should check this example: http://blog.justsophie.com/python-speech-to-text-with-pocketsphinx/
if any noise is detected above the ambient average, it start recording the audio until silence is detected for some time. I've been using it with some adaptation and it's working pretty well
@ankitj05 It should be possible. You should provide at least one hotword model (it can be any model), then look at the return results. If the return value is -2, then it means silence, otherwise it's non-silence.
Is there any way i can use only voice activity detection without implementing hotword detection?
I am using it in this way: ret = snowboydetect.SnowboyDetect.RunDetection(data) data is my frame. But its not working.