/// <summary>
/// Sends the voice audio to Google's API and runs HandleSpeech with transcription.
/// </summary>
private void TranscribeSpeech(Message m)
{
if (m.voice == null)
{
throw new Exception.EmptyVoiceMessageException(m);
}
if (m.voice.Duration > maxDur)
{
MaxDurationExceeded(m);
return;
}
SpeechClient speech = SpeechClient.Create();
RecognitionConfig config = new RecognitionConfig();
config.Encoding = SpeechHandler.VoiceTypeToGoogleType(m.voice.type);
config.SampleRateHertz = m.voice.sampleRate;
config.LanguageCode = languageCode;
config.ProfanityFilter = false;
RecognizeResponse resp = speech.Recognize(config, RecognitionAudio.FromStream(m.voice.AudioStream));
foreach (var result in resp.Results)
{
foreach (var alternative in result.Alternatives)
{
HandleSpeech(m, alternative.Transcript);
}
}
}
public static void UploadAudio(IFormFile audio)
{
// Reference to Google Cloud Speech-to_text Credentials
string credential_path = @"C:\Users\Billy\workspace\capstones\1000Words\1000Words\words-247918-f13fa4057b4a.json";
System.Environment.SetEnvironmentVariable("GOOGLE_APPLICATION_CREDENTIALS", credential_path);
using (Stream stream = audio.OpenReadStream())
{
RecognitionAudio recognitionAudio = RecognitionAudio.FromStream(stream);
var speech = SpeechClient.Create();
var response = speech.Recognize(new RecognitionConfig()
{
Encoding = RecognitionConfig.Types.AudioEncoding.Linear16,
LanguageCode = "en",
}, recognitionAudio);
Keywords.Clear();
foreach (var result in response.Results)
{
foreach (var alternative in result.Alternatives)
{
// Add transcript to list of keywords to be returned
Keywords.Add(alternative.Transcript);
}
}
}
}
private static RecognitionAudio LoadResourceAudio(string name)
{
var type = typeof(SpeechClientSnippets);
using (var stream = type.GetTypeInfo().Assembly.GetManifestResourceStream($"{type.Namespace}.{name}"))
{
return(RecognitionAudio.FromStream(stream));
}
}
private void OnComplete(Stream stream)
{
lock (this._lock)
{
var wavStream = RecognitionAudio.FromStream(stream);
var response = this._speech.Recognize(this._config, wavStream);
var alternatives =
from r in response.Results
from a in r.Alternatives
select new AlternativeBridge(a) as IAlternative;
this.RecognizeCompleted?.Invoke(alternatives.ToList());
}
}
// Not an actual test... just examples
public void FactoryMethods()
{
// Sample: FactoryMethods
RecognitionAudio audio1 = RecognitionAudio.FromFile("Sound/SpeechSample.flac");
RecognitionAudio audio2 = RecognitionAudio.FromUri("https://.../HostedSpeech.flac");
RecognitionAudio audio3 = RecognitionAudio.FromStorageUri("gs://my-bucket/my-file");
byte[] bytes = ReadAudioData(); // For example, from a database
RecognitionAudio audio4 = RecognitionAudio.FromBytes(bytes);
using (Stream stream = OpenAudioStream()) // Any regular .NET stream
{
RecognitionAudio audio5 = RecognitionAudio.FromStream(stream);
}
// End sample
}
protected override void OnStop(object sender, StoppedEventArgs e)
{
if (this._speech is null)
{
return;
}
var wavStream = RecognitionAudio.FromStream(e.WaveStream);
var results = this._speech.Recognize(this._config, wavStream).Results;
var alternatives = results.SelectMany(r => r.Alternatives);
Debug.WriteLine("\n========== Results ==========");
foreach (var alternative in alternatives)
{
Debug.WriteLine(alternative.Transcript + $"[{alternative.Confidence}]");
}
Debug.WriteLine("=============================\n");
this.Responsed?.Invoke(alternatives.ToList());
}
internal static string Parse(Stream stream)
{
var speech = SpeechClient.Create();
var response = speech.Recognize(new RecognitionConfig()
{
Encoding = RecognitionConfig.Types.AudioEncoding.Flac,
SampleRateHertz = 16000,
LanguageCode = "en",
}, RecognitionAudio.FromStream(stream));
string resultText = string.Empty;
foreach (var result in response.Results)
{
foreach (var alternative in result.Alternatives)
{
resultText = alternative.Transcript;
}
}
return(resultText);
}
public async Task <List <string> > ProcessSpeech(IFormFile speechBlob)
{
var speech = SpeechClient.Create();
var response = speech.Recognize(new RecognitionConfig()
{
Encoding = RecognitionConfig.Types.AudioEncoding.Linear16,
SampleRateHertz = 16000,
LanguageCode = "ro",
}, RecognitionAudio.FromStream(speechBlob.OpenReadStream()));
var transcripts = new List <string>();
foreach (var result in response.Results)
{
foreach (var alternative in result.Alternatives)
{
transcripts.Add(alternative.Transcript);
}
}
return(await Task.FromResult(transcripts));
}
public ActionResult <IEnumerable <string> > Post(IFormFile content)
{
List <string> returnResult = new List <string>();
if (content == null)
{
throw new Exception("Dosya Bulunamadı!");
}
try
{
Environment.SetEnvironmentVariable("GOOGLE_APPLICATION_CREDENTIALS", credential_path);
var speech = SpeechClient.Create();
if (content.Length > 0)
{
var response = speech.Recognize(new RecognitionConfig()
{
Encoding = RecognitionConfig.Types.AudioEncoding.Linear16,
SampleRateHertz = 8000,
LanguageCode = "tr-TR",
}, RecognitionAudio.FromStream(content.OpenReadStream()));
foreach (var result in response.Results)
{
foreach (var alternative in result.Alternatives)
{
returnResult.Add(alternative.Transcript);
}
}
}
return(returnResult);
}
catch (Exception e)
{
throw;
}
}
static void Main(string[] args)
{
String envName = "GOOGLE_APPLICATION_CREDENTIALS";
String envValue = System.IO.Path.GetDirectoryName(System.Reflection.Assembly.GetExecutingAssembly().GetName().CodeBase).Remove(0, 6) + "\\token.json";
//Stream stream = new FileStream(envValue, FileMode.Open, FileAccess.Read, FileShare.Read);
//var credential = GoogleCredential.FromStream(stream);
// Determine whether the environment variable exists.
//if (Environment.GetEnvironmentVariable(envName) == null)
// // If it doesn't exist, create it.
// Environment.SetEnvironmentVariable(envName, envValue);
//WaveInCapabilities deviceInfo = WaveIn.GetCapabilities(0);
Console.WriteLine("Gravando...");
WaveInEvent waveSource = new WaveInEvent();
//waveSource.DeviceNumber = 0;
waveSource.WaveFormat = new WaveFormat(16000, 1);
waveSource.DataAvailable += new EventHandler <WaveInEventArgs>(waveSource_DataAvailable);
MemoryStream tempFile = new MemoryStream();
waveFile = new WaveFileWriter(tempFile, waveSource.WaveFormat);
waveSource.StartRecording();
Console.WriteLine("Aperte Enter para parar de gravar e enviar para reconhecimento");
Console.ReadLine();
waveSource.StopRecording();
tempFile.Position = 0;
var speech = SpeechClient.Create();
//var response = speech.StreamingRecognize();
// Console.ReadLine();
long controlLen = tempFile.Length;
long position = 0;
DateTime dt = DateTime.Now;
//while ((DateTime.Now - dt).Minutes < 1)
{
position = tempFile.Length - controlLen;
controlLen = tempFile.Length;
tempFile.Position = position;
try
{
//Console.WriteLine("Reconhecendo...");
//var t = speech.AsyncRecognize(new RecognitionConfig()
//{
// Encoding = RecognitionConfig.Types.AudioEncoding.Linear16,
// SampleRate = 16000,
// LanguageCode = "pt-BR"
//}, RecognitionAudio.FromStream(tempFile));
//Console.WriteLine("{0}", t.PollUntilCompleted().Result);
Console.WriteLine("Reconhecendo...");
var response = speech.SyncRecognize(new RecognitionConfig()
{
Encoding = RecognitionConfig.Types.AudioEncoding.Linear16,
SampleRate = 16000,
LanguageCode = "pt-BR",
}, RecognitionAudio.FromStream(tempFile));
foreach (var result in response.Results)
{
foreach (var alternative in result.Alternatives)
{
Console.WriteLine(alternative.Transcript);
}
}
}
finally
{
Console.WriteLine("Reconhecido em {0}", DateTime.Now - dt);
dt = DateTime.Now;
}
}
waveFile.Dispose();
// Console.WriteLine("Reconhecimento Completo. Tempo={0}", DateTime.Now - dt);
Console.WriteLine("Terminou");
Console.ReadKey();
}