Overview

This exercise demonstrates how to translate speech using Azure AI Speech, enabling users to convert spoken language into translated text and synthesized speech.

Steps & Configuration Details

1. Provision an Azure AI Speech Resource

  • Open Azure Portal (https://portal.azure.com) and sign in.
  • Search for Azure AI services → Select Create under Speech service.
  • Configuration Items:
    • Subscription: Your Azure subscription.
    • Resource Group: Select or create a resource group.
    • Region: Choose any available region.
    • Name: Enter a unique name.
    • Pricing Tier: F0 (Free) or S (Standard).
    • Responsible AI Notice: Agree.

After provisioning, navigate to Keys and Endpoint in the Resource Management section.

2. Clone the Repository

  • Open Azure Cloud Shell in the Azure Portal.
  • Select PowerShell as the environment.
  • Run the following commands:
    rm -r mslearn-ai-language -f
git clone https://github.com/MicrosoftLearning/mslearn-ai-language mslearn-ai-language
cd mslearn-ai-language/Labfiles/08-speech-translation

3. Configure Your Application

  • Navigate to the correct folder:
    cd CSharp/translator  # For C#
cd Python/translator   # For Python
  • Install dependencies:
    • C#:
      dotnet add package Microsoft.CognitiveServices.Speech --version 1.30.0
    • Python:
      pip install azure-cognitiveservices-speech==1.30.0
  • Open the configuration file:
    • C#: appsettings.json
    • Python: .env
  • Update Configuration Values:
    • Azure AI Speech Region
    • API Key
  • Save the configuration file.

4. Add Code to Use Azure AI Speech SDK

  • Open the code file:
    • C#: Program.cs
    • Python: translator.py
  • Add references:
    • C#:
      using Microsoft.CognitiveServices.Speech;
using Microsoft.CognitiveServices.Speech.Audio;
using Microsoft.CognitiveServices.Speech.Translation;
    • Python:
      import azure.cognitiveservices.speech as speech_sdk
  • Configure the AI Speech client:
    • C#:
      translationConfig = SpeechTranslationConfig.FromSubscription(aiSvcKey, aiSvcRegion);
translationConfig.SpeechRecognitionLanguage = "en-US";
translationConfig.AddTargetLanguage("fr");
translationConfig.AddTargetLanguage("es");
translationConfig.AddTargetLanguage("hi");
Console.WriteLine("Ready to translate from " + translationConfig.SpeechRecognitionLanguage);
    • Python:
      translation_config = speech_sdk.translation.SpeechTranslationConfig(ai_key, ai_region)
translation_config.speech_recognition_language = 'en-US'
translation_config.add_target_language('fr')
translation_config.add_target_language('es')
translation_config.add_target_language('hi')
print('Ready to translate from', translation_config.speech_recognition_language)

5. Implement Speech Translation

  • C#:
    using AudioConfig audioConfig = AudioConfig.FromDefaultMicrophoneInput();
using TranslationRecognizer translator = new TranslationRecognizer(translationConfig, audioConfig);
Console.WriteLine("Speak now...");
TranslationRecognitionResult result = await translator.RecognizeOnceAsync();
Console.WriteLine($"Translating '{result.Text}'");
translation = result.Translations[targetLanguage];
Console.OutputEncoding = Encoding.UTF8;
Console.WriteLine(translation);
  • Python:
    audio_config = speech_sdk.AudioConfig(use_default_microphone=True)
translator = speech_sdk.translation.TranslationRecognizer(translation_config, audio_config=audio_config)
print("Speak now...")
result = translator.recognize_once_async().get()
print('Translating "{}"'.format(result.text))
translation = result.translations[targetLanguage]
print(translation)

6. Synthesize Translated Speech

  • C#:
    var voices = new Dictionary<string, string> {
    ["fr"] = "fr-FR-HenriNeural",
    ["es"] = "es-ES-ElviraNeural",
    ["hi"] = "hi-IN-MadhurNeural"
};
speechConfig.SpeechSynthesisVoiceName = voices[targetLanguage];
using SpeechSynthesizer speechSynthesizer = new SpeechSynthesizer(speechConfig);
SpeechSynthesisResult speak = await speechSynthesizer.SpeakTextAsync(translation);
if (speak.Reason != ResultReason.SynthesizingAudioCompleted) {
    Console.WriteLine(speak.Reason);
}
  • Python:
    voices = {
    "fr": "fr-FR-HenriNeural",
    "es": "es-ES-ElviraNeural",
    "hi": "hi-IN-MadhurNeural"
}
speech_config.speech_synthesis_voice_name = voices.get(targetLanguage)
speech_synthesizer = speech_sdk.SpeechSynthesizer(speech_config)
speak = speech_synthesizer.speak_text_async(translation).get()
if speak.reason != speech_sdk.ResultReason.SynthesizingAudioCompleted:
    print(speak.reason)

7. Run Your Application

  • C#:
    dotnet run
  • Python:
    python translator.py
  • Example prompt:
    Where is the station?
  • The response should display the transcribed speech, translated text, and synthesized speech output.

8. Clean Up

  • Delete Azure resources to avoid unnecessary costs:

Related Posts