西夏普的部落格

1.前言:
上一篇文章介紹text-to-speech，這篇範例介紹語音辨識方法，不過電腦需要長時間的訓練辨識率才能提高。

2.說明:
有關Microsoft Speech API的說明請參考MSDN:
http://msdn.microsoft.com/en-us/library/ee125663(v=vs.85).aspx

加入參考
在COM元件中選擇Microsoft Speech Object Library 5.4

加入命名空間:

using SpeechLib;

程式碼:

private void button3_Click(object sender, EventArgs e)
{
	SpSharedRecoContextClass objRecoContext = null; ;
	ISpeechRecoGrammar grammar;
	objRecoContext = new SpSharedRecoContextClass();
	objRecoContext.Recognition += new _ISpeechRecoContextEvents_RecognitionEventHandler(objRecoContext_Recognition);
	grammar = objRecoContext.CreateGrammar(0);
	grammar.DictationLoad("", SpeechLoadOption.SLOStatic);
	grammar.DictationSetState(SpeechRuleState.SGDSActive);
}

public void objRecoContext_Recognition(int StreamNumber, object StreamPosition, SpeechRecognitionType RecognitionType, ISpeechRecoResult Result)
{
	string recgData = Result.PhraseInfo.GetText(0, -1, true);
	textBox1.Text = recgData;
	Speak(recgData);
	recgData = "";
}