Translator Service

A lightweight library that uses Cognitive Translator Service for text translation and Cognitive Speech Service for text-to-speech and speech-to-text.

To use this library, you must register an All-In-One Cognitive Service or Translator Service and Speech Service on Azure to obtain the Subscription keys. Keep in mind that Regional endpoints are available, so you need to pass the region name to the library according to your registration.

Installation

The library is available on NuGet. Just search TranslatorService in the Package Manager GUI or run the following command in the Package Manager Console:   

Install-Package TranslatorService

Usage

Translation

It's usage is straightforward. For example, if you want to translate text:

// Use a Global Translator Service
var translatorClient = new TranslatorClient("<subscription_key>");

// Use a Regional Translation Service
var translatorClient = new TranslatorClient("<subscription_key>", "<region>");

// Use an external HttpClient (i.e. coming from IHttpClientFactory)
var translatorClient = new TranslatorClient(httpClient, "<subscription_key>", "<region>");

var response = await translatorClient.TranslateAsync("Today is really a nice day.", to: "it");
Console.WriteLine(
 $"Detected source language: {response.DetectedLanguage.Language} ({response.DetectedLanguage.Score:P2})");

Console.WriteLine(response.Translation.Text);

Speech Recognition

var speechClient = new SpeechClient("<subscription_key>", "<region>");

// Use an external HttpClient (i.e. coming from IHttpClientFactory)
var speechClient = new SpeechClient(httpClient, "<subscription_key>", "<region>");

var response = await speechClient.RecognizeAsync(audioStream, "en-US", RecognitionResultFormat.Detailed);
Console.WriteLine($"Recognition Result: {response.RecognitionStatus}");
Console.WriteLine(response.DisplayText);

Speech Synthesis

var speechClient = new SpeechClient("<subscription_key>", "<region>");

// Use an external HttpClient (i.e. coming from IHttpClientFactory)
var speechClient = new SpeechClient(httpClient, "<subscription_key>", "<region>");

var responseStream = await speechClient.SpeakAsync(new TextToSpeechParameters
    {
        Language = "en-US",
        VoiceType = Gender.Female,
        VoiceName = "en-US-AriaNeural",
        Text = "Hello everyone! Today is really a nice day.",
    });

In the Samples folder are available samples for .NET Core and the Universal Windows Platform.

Contribute

The project is continually evolving. We welcome contributions. Feel free to file issues and pull requests on the repo and we'll address them as we can.