detect speaking events from web audio
npm install @rifflearning/sibilant
Sibilant emits three different kinds of events after it's been hooked up to a web audio stream:
speaking
: emitted continuously while voiced speech is detectedstoppedSpeaking
: emitted at the end of an utterance, with the start and end time of speechvolumeChanged
: emitted any time the volume of the audio stream changes
import Sibilant from '@rifflearning/sibilant';
var speakingEvents = new Sibilant(stream)
speakingEvents.bind('speaking', function () {
console.log('speaking!')
})
speakingEvents.bind('stoppedSpeaking', function (data) {
console.log('Spoke!', 'start:', data.start, 'end:', data.end)
})
speakingEvents.bind('volumeChanged', function (data) {
console.log('volume changed!', 'value:', data)
})
MIT