This repository is deprecated!

---

adaptive-speech v0.3.0

This module allows you to control web app using voice commands. It’s based on Chrome’s speech recognition API.

Demo

Check out http://angular-adaptive.github.io/adaptive-speech/demo/

References

We recomend you to read:

• A More Awesome Web presentation from Google IO 2013 by Eric Bidelman
• Voice Driven Web Apps article from HTML5 ROCKS by Glen Shires

Requirements

• AngularJS v ~1.2.x

Usage

We use bower for dependency management. Add

dependencies: {
    "angular-adaptive-speech": "latest"
}

To your bower.json file. Then run

bower install

This will copy the speech recognition files into your bower_components folder, along with its dependencies. Load the script files in your application:

<script type="text/javascript" src="bower_components/angular/angular.js"></script>
<script type="text/javascript" src="bower_components/angular-adaptive-speech/angular-adaptive-speech.min.js"></script>

Add the adaptive.speech module as a dependency to your application module:

var myAppModule = angular.module('MyApp', ['adaptive.speech']);

and include $speechRecognition, $speechSynthetis, $speechCorrection service as a dependency to your controller:

angular.module('MyApp').controller('MainCtrl', function ['$scope', '$speechRecognition, $speechSynthetis', ($scope, $speechRecognition, $speechSynthetis) {

}]);

To start speech recognition run from controller:

$speechRecognition.onstart(function(){
  $speechSynthetis.speak('Yes? How can I help you?', 'en-UK');
});
$speechRecognition.setLang('en-UK'); // Default value is en-US
$speechRecognition.listen();

Apply the directive to your elements where reference is keyword reference:

<ul>
    <li ng-repeat="todo in todos | filter:statusFilter track by $index" speechrecognition="{'tasks': recognition['en-US']['listTasks'], 'reference': todo}">
        {{todo}}
    </li>
</ul>

Or run recognition directly from controller:

$speechRecognition.listenUtterance($scope.recognition['en-US']['addToList']);

Options

All the speechRecognition options can be set up in your controller.

myAppModule.controller('MyController', function($scope) {
    $scope.recognition = {};
    $scope.recognition['en-US'] = {
        'addToList': {
            'regex': /^to do .+/gi,
            'lang': 'en-US',
            'call': function(e){
                $scope.addToList(e);
            }
        },
        'listTasks': [{
            'regex': /^complete .+/gi,
            'lang': 'en-US',
            'call': function(e){
                $scope.completeTask(e);
            }
        },{
            'regex': /^remove .+/gi,
            'lang': 'en-US',
            'call': function(e){
                $scope.removeTask(e);
            }
        }]
    };
});

APIs

Check out API docs.

Testing

We use karma and jshint to ensure the quality of the code. The easiest way to run these checks is to use grunt:

npm install -g grunt-cli
npm install
bower install
grunt

The karma task will try to open Chrome as a browser in which to run the tests. Make sure this is available or change the configuration in test/test.config.js

Contributing

Pull requests are welcome.

Make a PR against canary branch and don’t bump any versions.

Please respect the code style in place.

License

The MIT License

Copyright © 2014 Jan Antala