getusermedia-to-text
getUserMedia to Text via Google's Speech to Text API.
npm install getusermedia-to-text --save
Usage
var GetUserMediaToText = var path = var s2t = projectId: 'your-project-id' keyFilename: path // request: { options } s2ts2ts2ts2t s2tstarts2t
Unfortunately google's speech API module depends on grpc which is a native module. Native modules don't work in the browser so this probably only works for electron after you run electron-rebuild.
To run the example:
npm install
npm start
You are required to set up your own service account since the speech recognition API requires an account with billing. See:
https://cloud.google.com/speech/docs/common/auth
The example code expects to find your service account credentials at ./example/auth.json
.
API
var s2t = new GetUserMediaToText(options)
Create a new speech to text instance.
Options include:
keyFilename
(required): path to your auth.json file.projectId
: your project id string. Default:getusermedia-to-text
request
: the request options object. Default:
config: encoding: 'LINEAR16' sampleRateHertz: 44100 languageCode: 'en-US' singleUtterance: false interimResults: false verbose: true
s2t.start()
Start listening to audio. Will event speech events as data
.
s2t.stop()
Stop listening to audio and destroy internal recognition stream.
s2t
events:
error
: Emitted when an error is encountered with the error object.status
: General status message strings.listening
: Emits boolean listening status data when the listening status changes.data
: Raw data objects that are emitted from the@google-cloud/speech
library.