ms-speech

Note about this repository

This is a fork of the microsoft-speech-browser-sdk microsoft-speech-browser-sdk package on npm. This version is built using CommonJS modules, making it suitable for use with codebases that use either CommonJS or ES6 module import syntax, particularly when used with webpack.

The text below is from the original README. For the official version of this SDK maintained by Microsoft, see https://github.com/Azure-Samples/SpeechToText-WebSockets-Javascript.

Background

Microsoft's Speech Service is a cloud-based platform that features the most advanced algorithms available for converting spoken audio to text. The Universal Speech Protocol allows you to integrate speech recognition into your application using the Microsoft Speech Service.

Install

To install npm package run

npm install microsoft-speech-browser-sdk

JavaScript SDK - Sample Usage

RequireJs is a dependency. Make sure to reference it in your page before using the SDK.

// Resolve the SDK dependecy using RequireJs
require(["Speech.Browser.Sdk"], function(SDK) {
    // Now start using the SDK
});
 
// Setup the recongizer
function RecognizerSetup(SDK, recognitionMode, language, format, subscriptionKey) {
    let recognizerConfig = new SDK.RecognizerConfig(
        new SDK.SpeechConfig(
            new SDK.Context(
                new SDK.OS(navigator.userAgent, "Browser", null),
                new SDK.Device("SpeechSample", "SpeechSample", "1.0.00000"))),
        recognitionMode, // SDK.RecognitionMode.Interactive  (Options - Interactive/Conversation/Dictation)
        language, // Supported laguages are specific to each recognition mode. Refer to docs.
        format); // SDK.SpeechResultFormat.Simple (Options - Simple/Detailed)
 
    // Alternatively use SDK.CognitiveTokenAuthentication(fetchCallback, fetchOnExpiryCallback) for token auth
    let authentication = new SDK.CognitiveSubscriptionKeyAuthentication(subscriptionKey);
 
    return SDK.Recognizer.Create(recognizerConfig, authentication);
}
 
function RecognizerStart(SDK, recognizer) {
    recognizer.Recognize((event) => {
        /*
            Alternative syntax for typescript devs.
            if (event instanceof SDK.RecognitionTriggeredEvent)
        */
        switch (event.Name) {
            case "RecognitionTriggeredEvent" :
                UpdateStatus("Initializing");
                break;
            case "ListeningStartedEvent" :
                UpdateStatus("Listening");
                break;
            case "RecognitionStartedEvent" :
                UpdateStatus("Listening_Recognizing");
                break;
            case "SpeechStartDetectedEvent" :
                UpdateStatus("Listening_DetectedSpeech_Recognizing");
                console.log(JSON.stringify(event.Result)); // check console for other information in result
                break;
            case "SpeechHypothesisEvent" :
                UpdateRecognizedHypothesis(event.Result.Text);
                console.log(JSON.stringify(event.Result)); // check console for other information in result
                break;
            case "SpeechFragmentEvent" :
                UpdateRecognizedHypothesis(event.Result.Text);
                console.log(JSON.stringify(event.Result)); // check console for other information in result
                break;
            case "SpeechEndDetectedEvent" :
                OnSpeechEndDetected();
                UpdateStatus("Processing_Adding_Final_Touches");
                console.log(JSON.stringify(event.Result)); // check console for other information in result
                break;
            case "SpeechSimplePhraseEvent" :
                UpdateRecognizedPhrase(JSON.stringify(event.Result, null, 3));
                break;
            case "SpeechDetailedPhraseEvent" :
                UpdateRecognizedPhrase(JSON.stringify(event.Result, null, 3));
                break;
            case "RecognitionEndedEvent" :
                OnComplete();
                UpdateStatus("Idle");
                console.log(JSON.stringify(event)); // Debug information
                break;
        }
    })
    .On(() => {
        // The request succeeded. Nothing to do here.
    },
    (error) => {
        console.error(error);
    });
}
 
function RecognizerStop(SDK, recognizer) {
    // recognizer.AudioSource.Detach(audioNodeId) can be also used here. (audioNodeId is part of ListeningStartedEvent)
    recognizer.AudioSource.TurnOff();
}

Try the sample out

What to try the sample ? All you need is a subscription key. Sign up to get one.

Here is a handy link to our Sample that you can try out. (Rendered using htmlPreview)

Note: Some browsers block microphone access on un-secure origin. So, it is recommended to host the 'sample'/'your app' on https to get it working on all supported browsers.

Docs

The sdk is a reference implementation for the speech websocket protocol. Check the API reference and Websocket protocol reference for more details.

Browser support

The SDK depends on WebRTC apis to get access to the microphone and read the audio stream. Most of todays browsers(Edge/Chrome/Firefox) support this. For more details about supported browsers refer to navigator.getUserMedia#BrowserCompatibility

Note: The sdk currently depends on navigator.getUserMedia api. However this api is in process of being dropped as browsers are moving towards newer MediaDevices.getUserMedia instead. The sdk will add support to the newer api soon.

Contributing

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact [e-mail] with any additional questions or comments.