static void StartRecognition()
{
var audioDirectory = Path.Combine(Directory.GetCurrentDirectory(), "Audio");
var audioCollection = new Dictionary <string, FileInfo>();
foreach (var item in Directory.GetFiles(audioDirectory))
{
var audioFile = new FileInfo(item);
audioCollection.Add(audioFile.Name, audioFile);
}
var audioListString = string.Join(" \n", audioCollection.Keys);
Console.WriteLine(@"Which file to use -> {0}", audioListString);
var response = Console.ReadLine();
if (string.IsNullOrEmpty(response))
{
response = "robot.wav";
}
_recognizer.StartRecognition(new FileStream(Path.Combine(audioDirectory, response), FileMode.Open));
Console.WriteLine(@"Press any key to start Speech Recognition...");
Console.ReadLine();
var result = _recognizer.GetResult();
_recognizer.StopRecognition();
if (result != null)
{
Console.WriteLine(result.GetHypothesis());
}
}
public void Transcribe_UsingGrammar_Continuous()
{
_configuration = new Configuration
{
AcousticModelPath = ModelsDirectory,
DictionaryPath = _dictionaryPath,
LanguageModelPath = _languageModelPath,
UseGrammar = true,
GrammarName = "hello",
GrammarPath = ModelsDirectory
};
_speechRecognizer = new StreamSpeechRecognizer(_configuration);
for (int i = 1; i <= 3; i++)//TODO: Reading 3 or more files in a row causes test fail (The same happens in CMU Sphinx4)
{
var audioFile = Path.Combine(Directory.GetCurrentDirectory(), "Audio", string.Format("Long Audio {0}.wav", i));
var stream = new FileStream(audioFile, FileMode.Open);
_speechRecognizer.StartRecognition(stream);
var result = _speechRecognizer.GetResult();
_speechRecognizer.StopRecognition();
Assert.IsNotNull(result);
var hypothesis = result.GetHypothesis();
Assert.IsTrue(hypothesis.Contains("the time is now exactly twenty five to one") || hypothesis.Contains("there's three left on the left side the one closest to us"));
}
}
public async Task <string> ProcessVoiceToText(Stream stream, int bitRate)
{
var fn = $"a-{Guid.NewGuid()}-{bitRate}.wav";
stream.Seek(0, SeekOrigin.Begin);
var wavStream = new RawSourceWaveStream(stream, new WaveFormat(bitRate, 2));
// Debugging only
// WaveFileWriter.CreateWaveFile($"{fn}-source.wav", wavStream);
stream.Seek(0, SeekOrigin.Begin);
var newFormat = new WaveFormat(InputRate, 1);
WaveFormatConversionStream cs = new WaveFormatConversionStream(newFormat, wavStream);
// Debugging only
// WaveFileWriter.CreateWaveFile(fn, cs);
cs.Seek(0, SeekOrigin.Begin);
speechRecognizer.StartRecognition(cs);
var result = speechRecognizer.GetResult();
speechRecognizer.StopRecognition();
cs.Close();
return(result?.GetHypothesis());
}
public void StartRecognition(Stream stream)
{
try
{
CanTranscribe = false;
_recognizer.StartRecognition(stream);
var result = _recognizer.GetResult();
_recognizer.StopRecognition();
if (result != null)
{
MessageBox.Show(result.GetHypothesis());
}
stream.Close();
CanTranscribe = true;
}
catch (Exception exception)
{
this.LogError(exception);
}
}
private void recognize()
{
waveInStream.StopRecording();
mem.Position = 0;
_recognizer.StartRecognition(mem, new TimeFrame(mem.Length));
SpeechResult result = _recognizer.GetResult();
_recognizer.StopRecognition();
Console.WriteLine("result: " + result.GetHypothesis());
Console.ReadKey();
Console.ReadKey();
}
private static void Main()
{
Logger.LogReceived += LogReceived;
var modelsDirectory = Path.Combine(Directory.GetCurrentDirectory(), "Models");
var audioDirectory = Path.Combine(Directory.GetCurrentDirectory(), "Audio");
if (!Directory.Exists(modelsDirectory) || !Directory.Exists(audioDirectory))
{
Console.WriteLine("No Models or Audio directory found!! Aborting...");
Console.ReadLine();
return;
}
_speechConfiguration = new Configuration
{
AcousticModelPath = modelsDirectory,
DictionaryPath = Path.Combine(modelsDirectory, "cmudict-en-us.dict"),
LanguageModelPath = Path.Combine(modelsDirectory, "en-us.lm.dmp"),
UseGrammar = true,
GrammarPath = modelsDirectory,
GrammarName = "hello"
};
_speechRecognizer = new StreamSpeechRecognizer(_speechConfiguration);
for (int i = 1; i <= 3; i++)
{
var audioFile = Path.Combine(audioDirectory, String.Format("Long Audio {0}.wav", i));
var stream = new FileStream(audioFile, FileMode.Open);
if (i == 3)
{
System.Diagnostics.Trace.WriteLine("checking");
}
_speechRecognizer.StartRecognition(stream);
var result = _speechRecognizer.GetResult();
_speechRecognizer.StopRecognition();
if (result != null)
{
Console.WriteLine(string.Format("Result: {0}", i) + result.GetHypothesis());
}
else
{
Console.WriteLine("Result: {0}", "Sorry! Coudn't Transcribe");
}
var instance = ConfigurationManager.GetInstance <SpeechMarker>();
Console.WriteLine(instance.ToString());
stream.Close();
}
Console.WriteLine("DONE!");
Console.ReadLine();
}
