Windows 对话的语音识别

Question

如何在 UWP 应用程序中使用 Windows 语音识别进行对话？目前，它只能识别我的声音，不能识别同一对话中其他说话者的声音。也许，是否存在另一个 API？

这是我的原始代码：

public sealed partial class MainPage : Page
    {
        private SpeechRecognizer speechRecognizer;

        public MainPage()
        {
            this.InitializeComponent();
            this.speechRecognizer = new SpeechRecognizer();

            Init();
        }

        private async void Init()
        {
            //Compile predifined grammar
            SpeechRecognitionCompilationResult result = await speechRecognizer.CompileConstraintsAsync();

            speechRecognizer.ContinuousRecognitionSession.ResultGenerated += ContinuousRecognitionSession_ResultGenerated;

            if (speechRecognizer.State == SpeechRecognizerState.Idle)
            {
                await speechRecognizer.ContinuousRecognitionSession.StartAsync();
            }
        }

        private async void ContinuousRecognitionSession_ResultGenerated(SpeechContinuousRecognitionSession sender,
                                                                        SpeechContinuousRecognitionResultGeneratedEventArgs args)
        {
            Console.WriteLine(args.Result.Text);
        }
}

Answer 1

SpeechRecognizer 不专门识别人的声音。它会分析当前输入的声音，并通过相应的语法将其转换为文本输出。

所以SpeechRecognizer最理想的对话场景是两个人离麦克风不远，说话清晰，同时只有一个人在说话。

SpeechRecognizer提供的是简单的语音识别服务，没有提供API分离两个人的声音并分别识别输出。

Windows 对话的语音识别

Windows Speech Recognition for conversations

c#

windows

speech-recognition

uwp