Question

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
using System.Speech.Recognition;

namespace ConsoleApp2
{
  class Program
  {
     static SpeechRecognitionEngine recEngine = new 
     SpeechRecognitionEngine();
     bool keyHold = false;

     static void Main(string[] args)
     {
        Choices commands = new Choices();
        commands.Add(new string[] { "Current dollar value", "Current euro value" });
        GrammarBuilder gBuilder = new GrammarBuilder();
        gBuilder.Append(commands);
        Grammar grammar = new Grammar(gBuilder);

        recEngine.LoadGrammarAsync(grammar);
        recEngine.SetInputToDefaultAudioDevice();
        recEngine.RecognizeAsync(RecognizeMode.Multiple);
     }

     void recEngine_SpeechRecognized(object sender, SpeechRecognizedEventArgs e)
     {
         switch (e.Result.Text)
         {
            case "Current dollar value":
                Console.WriteLine("10kr");
                break;

            case "Current euro value":
                Console.WriteLine();
                break;
        }
     }
  }
}

该程序刚刚启动时关闭，并且没有录制我的声音，我尝试在代码console.readkey();中的recEngine.RecognizeAsync(RecognizeMode.Multiple);下放置此行。它保持程序并且没有自动关闭，但程序仍然不记录我的传入语音命令这是为什么？

Answer 1

语音识别是事件驱动的您需要定义必要的事件处理程序必须订阅SpeechRecognitionEngine.SpeechRecognized Even才能收到认可结果另见SpeechRecognitionEngine.AudioStateChanged Event

此外，在开始识别之前，必须使用新设置更新引擎：
见SpeechRecognitionEngine.RequestRecognizerUpdate Method

注意：
我只使用＆＃34; 美元＆＃34;和＆＃34; 欧元＆＃34;作为语法项目的提示：尝试在你的语法中使用可能较少的关键词，否则他们变得非常难以处理。您可以发出整个＆＃34;当前美元价值＆＃34;短语，然而，公认的关键词是＆＃34;美元＆＃34;。

using System.Speech.Recognition;

  static void Main(string[] args)
  {
    using (SpeechRecognitionEngine recEngine = new SpeechRecognitionEngine())
    {
       recEngine.SpeechRecognized += 
             new EventHandler<SpeechRecognizedEventArgs>(recEngine_SpeechRecognized);
       recEngine.AudioStateChanged += 
             new EventHandler<AudioStateChangedEventArgs>(recEngine_AudioStateChange);

       Choices commands = new Choices();
       commands.Add(new string[] { "dollar", "euro" });
       GrammarBuilder gBuilder = new GrammarBuilder();
       gBuilder.Append(commands);
       Grammar grammar = new Grammar(gBuilder);

       recEngine.SetInputToDefaultAudioDevice();
       recEngine.LoadGrammarAsync(grammar);
       recEngine.RequestRecognizerUpdate();

       recEngine.RecognizeAsync(RecognizeMode.Multiple);

       Console.ReadKey();
       recEngine.SpeechRecognized -= recEngine_SpeechRecognized;
       recEngine.AudioStateChanged -= recEngine_AudioStateChange;
    }
  }

  internal static void recEngine_AudioStateChange(object sender, AudioStateChangedEventArgs e)
  {
     Console.WriteLine("Current AudioLevel: {0}", e.AudioState);
  }

  internal static void recEngine_SpeechRecognized(object sender, SpeechRecognizedEventArgs e)
  {
     switch (e.Result.Text)
     {
        case "dollar":
           Console.WriteLine("10kr");
           break;
        case "euro":
           Console.WriteLine("A lot more");
           break;
     }
  }

注2：
在Windows 7上，System.Speech.Recognition无效。
可以使用Microsoft Speech Platform的免费开发工具启用语音识别和TTS（文本到语音）。下载并安装SDK，Runtime和Desidered语言。这是服务器语音识别实现，开发用于通过电话线支持语音识别。因此它具有非常好的原生滤波器，可以发出噪音，并且对发音怪癖具有良好的容忍度。

语音识别程序不捕获语音/命令

1 个答案: