Speech recognition in node and the browser using Electron.

It seems that Google has shut down the Chrome Speech API for use in shell environments like Electron, which electron-speech relies on.

Google does offer a general speech api, which you can access using a module like google-speech-api. Unfortunately, you will require ffmpeg, a Google API key, and internet connectivity.

Intrepid developers! A purely local, open, offline-friendly solution like jasper may be the road forward! Go forth and write some awesome node modules around it!

If you want to use the API from a script using Node, you will need to launch your script using the electron-spawn command instead of the node command:

$ npm install -g electron-spawn

$ electron-spawn example.js

This module installs the electron-speech command:

$ electron-speech
listening..
(whatever is said is written here, to stdout)

-q|--quiet to omit the listening.. message. (it's on stderr anyways though)

-c|--continuous to keep on listening after each result.

var Speech = require('electron-speech')

var recog = Speech({
  lang: 'en-US',
  continuous: true
})

recog.on('text', function (text) {
  console.log(text)
});

recog.listen()

Returns speech, an EventEmitter and Readable stream.

opts accepts multiple keys:

• opts.lang - recognize speech in the language lang. Defaults to 'en-US'.
• opts.continuous - if true, text events will keep on being emitted as recognition occurs.

Starts listening to speech via the microphone. 'ready'` will be emitted once speech recognition has begun.

Uses speech as a readable stream for text rather than an event emitter. Results have newlines appended to them for parsing convenience.

Emitted when the microphone has begun to listen for speech.

Emitted when speech has been recognized.

Emitted when an error has occurred in recognition.

Emitted when recognition has ended. Does not fire if continuous was set to true.

ISC