js-speech-rekognition

HTLM5 based speech recognition

Usage

Usage:

import {Rekognition} from 'js-speech-rekognition';

const r = new Rekognition();
r.setParams({
  lang: 'en-US',
  interimResults: false,
  continuous: false,
  matchThreshold: 0.8,
  grammar: [{
    phrase_id: 1,
    phrase: 'hello',
    aliases: ['hi there', 'wassup']
  },{
    phrase_id: 2,
    phrase: 'green',
    aliases: ['color']
  }, {
    phrase_id: 3,
    phrase: 'blue',
    aliases: ['color']
  }, {
    phrase_id: 4,
    phrase: 'red',
    aliases: ['color']
  }, {
    phrase_id: 5,
    phrase: 'yellow',
    aliases: ['color']
  }]
});

To enable debugging call:

r.debugOn();

To disable debugging call:

r.debugOff();

The service supports for the following custom events:

1. match - When the spoken command matches with the phrases in the grammar
2. error - When there's an error with the SpeechAPI
3. start - When speech rekognition service starts
4. stop - When speech rekognition service stops
5. partial_speech - When partial speech is detected
6. complete_speech - When final speech results are received

Attach event listeners as follow:

  r.on('match', e => {
    // do something
  });
  r.on('error', e => {
    // error occured
  });

To start the recognition service:

r.startRekognition();

To stop the recognition service:

r.stopRekognition();

To remove all listeners:

  r.removeListeners();

To reset all commands:

r.reset();