C#.NET语音识别应用遇System.UnauthorizedAccessException权限错误求助

Hey there, let's work through this System.UnauthorizedAccessException (E_ACCESSDENIED) issue you're hitting with your speech recognition project. I've looked over your error stack trace and code, and here are the most likely fixes to get your app running smoothly:

1. Verify Microphone Permissions on Windows #

First and foremost, your app needs explicit permission to access your system's microphone. Here's how to check:

• Open Windows Settings → Privacy & Security → Microphone
• Ensure the toggle for "Allow apps to access your microphone" is turned on
• If you're debugging in Visual Studio, make sure "Allow desktop apps to access your microphone" is also enabled (this covers WinForms apps)

2. Fix Concurrent Audio Device Access in Your Code #

Looking at your Form1_Load method, you're initializing two SpeechRecognitionEngine instances (speechRecognitionEngine and speechListener) and both are trying to grab the default audio device at the same time. This is a common cause of access denied errors—two processes can't hold exclusive access to the microphone simultaneously.

You also have redundant code: you're loading the same grammar twice into speechRecognitionEngine, and accidentally loading it into the wrong engine for speechListener. Here's a cleaned-up version of your Form1_Load method to avoid these issues:

private void Form1_Load(object sender, EventArgs e)
{
    // Initialize main recognition engine
    speechRecognitionEngine.SetInputToDefaultAudioDevice();
    var commandGrammar = new Grammar(new GrammarBuilder(new Choices(File.ReadAllLines(@"Commands.txt"))));
    speechRecognitionEngine.LoadGrammarAsync(commandGrammar);
    speechRecognitionEngine.SpeechRecognized += Default_SpeechRecognized;
    speechRecognitionEngine.SpeechRecognized += _recognizer_SpeechRecognized;
    speechRecognitionEngine.RecognizeAsync(RecognizeMode.Multiple);

    // Initialize listener engine but don't claim the mic yet
    var listenerGrammar = new Grammar(new GrammarBuilder(new Choices(File.ReadAllLines(@"Commands.txt"))));
    speechListener.LoadGrammarAsync(listenerGrammar);
    speechListener.SpeechRecognized += startlistening_SpeechRecognized;
    // Wait to set input and start listening until the user triggers "Stop listening"
}

Then update your command handlers to safely switch between engines and release the microphone when not in use:

Update "Stop listening" handler: #

if(speech == "Stop listening") {
    speechSynthesizer.SpeakAsync("Call me any time you want");
    speechRecognitionEngine.RecognizeAsyncCancel();
    speechRecognitionEngine.SetInputToNull(); // Release the mic from main engine
    speechListener.SetInputToDefaultAudioDevice(); // Claim mic for listener
    speechListener.RecognizeAsync(RecognizeMode.Multiple);
}

Update "Hello assistant" handler: #

private void startlistening_SpeechRecognized(object sender, SpeechRecognizedEventArgs e) {
    string speech = e.Result.Text;
    if(speech == "Hello assistant") {
        speechListener.RecognizeAsyncCancel();
        speechListener.SetInputToNull(); // Release mic from listener
        speechSynthesizer.SpeakAsync("Hello, what do you want");
        speechRecognitionEngine.SetInputToDefaultAudioDevice(); // Reclaim mic for main engine
        speechRecognitionEngine.RecognizeAsync(RecognizeMode.Multiple);
    }
}

3. Run Your App with Administrator Privileges #

Sometimes system-level audio access requires elevated permissions. Try:

• Right-clicking Visual Studio (or your compiled EXE) and selecting Run as administrator
• If debugging, you can set this permanently in Visual Studio: Right-click your project → Properties → Debug → Check "Enable native code debugging" (or just run VS as admin each time)

4. Check for Other Apps Using the Microphone #

Make sure no other applications (like Zoom, Teams, Discord, or recording software) are currently using your microphone. These apps can lock the device, preventing your C# app from accessing it.

内容的提问来源于stack exchange，提问作者SooryaSRajan