sl-web-speech
This is a library for handling speech on web with a focus on covering interactive storytelling use cases. In particular, continuous realtime speech recognition that aims to match expected keywords.
Features
- Stream audio from microphone to speech recognition (vosk-browser)
- Receive callbacks for recognized words, start of speech, and stop of speech
- Recognized word events are deduped
- Mute and unmute the microphone
- State management around waiting for browser permissions and enablements
Licensing
My code is licensed with the MIT open source license.
But this project also contains a "models" directory which is separately licensed under Apache. See COPYING file in that directory and follow its terms before redistributing.
Usage
Your web app will use 'sl-web-speech' as a dependency library. At time of writing, I've not published to NPM, so you'd need to git clone and use npm link or other solution to import into your web app. If anybody asks me to publish this on NPM, I'm willing to do that.
The files under "models" should be served from "/models" on your web server.
Example web app code is shown below.
Listen for Speech
import Recognizer from 'sl-web-speech';
/* Construct Recognizer *after* the user has performed some UI interaction in your web app. For security reasons, most
web browsers won't access microphone audio until there is a UI interaction happens. An easy way to accomplish this
is to have a starting page that requires a button/link click to begin listening for speech. */
const recognizer = new Recognizer(onReady);
function onReady() {
bindCallbacks(onPartial, onStartSpeaking, onStopSpeaking);
recognizer.unmute();
}
function onPartial(speechText) {
console.log(`User said "${speechText}".`);
}
function onStartSpeaking() {
console.log('User started speaking.');
}
function onStopSpeaking() {
console.log('User stopped speaking.');
}Contributing
The project isn't open to contributions at this point. But that could change. Contact me if you'd like to collaborate.
Contacting
You can reach me on LinkedIn. I'll accept connections if you will just mention "SL Web Speech" or some other shared interest in your connection request.