IBM Watson语音转文本服务未在Unity3d中给出响应 [英] IBM Watson Speech to Text Service is not giving response in Unity3d

问题描述

我有一个ExampleSstreaming类，实际上是从 IBM Watson SDK的GitHub获得的(语音文字服务演示).这是

I have an ExampleSstreaming class which actually I got from GitHub of IBM Watson SDK (speech to text service demo). Here it is

public class ExampleStreaming : MonoBehaviour
{
  private int m_RecordingRoutine = 0;
  private string m_MicrophoneID = null;
  private AudioClip m_Recording = null;
  private int m_RecordingBufferSize = 5;
  private int m_RecordingHZ = 22050;

  private SpeechToText m_SpeechToText = new SpeechToText();

  void Start()
  {
    LogSystem.InstallDefaultReactors();
    Log.Debug("ExampleStreaming", "Start();");

    Active = true;
        Debug.Log("start");
    StartRecording();
  }

    public void Update() {
        Debug.Log(m_SpeechToText.IsListening);
    }

  public bool Active
  {
    get { return m_SpeechToText.IsListening; }
    set
    {
      if (value && !m_SpeechToText.IsListening)
      {
        m_SpeechToText.DetectSilence = true;
        m_SpeechToText.EnableWordConfidence = false;
        m_SpeechToText.EnableTimestamps = false;
        m_SpeechToText.SilenceThreshold = 0.03f;
        m_SpeechToText.MaxAlternatives = 1;
        m_SpeechToText.EnableContinousRecognition = true;
        m_SpeechToText.EnableInterimResults = true;
        m_SpeechToText.OnError = OnError;
        m_SpeechToText.StartListening(OnRecognize);
      }
      else if (!value && m_SpeechToText.IsListening)
      {
        m_SpeechToText.StopListening();
      }
    }
  }

  private void StartRecording()
  {
    if (m_RecordingRoutine == 0)
    {
            Debug.Log("m_RecordingRoutine");
            UnityObjectUtil.StartDestroyQueue();
      m_RecordingRoutine = Runnable.Run(RecordingHandler());
    }
  }

  private void StopRecording()
  {
    if (m_RecordingRoutine != 0)
    {
      Microphone.End(m_MicrophoneID);
      Runnable.Stop(m_RecordingRoutine);
      m_RecordingRoutine = 0;
    }
  }

  private void OnError(string error)
  {
    Active = false;

    Log.Debug("ExampleStreaming", "Error! {0}", error);
  }

  private IEnumerator RecordingHandler()
  {
    Log.Debug("ExampleStreaming", "devices: {0}", Microphone.devices);

        m_MicrophoneID = Microphone.devices[0];
        Debug.Log("m_MicrophoneID : " + m_MicrophoneID);
        m_Recording = Microphone.Start(m_MicrophoneID, true, m_RecordingBufferSize, m_RecordingHZ);
    yield return null;      // let m_RecordingRoutine get set..
        Debug.Log("m_Recording : " + m_Recording.length);
        if (m_Recording == null)
    {
            Debug.Log("m_Recording is null");
            StopRecording();
      yield break;
    }

    bool bFirstBlock = true;
    int midPoint = m_Recording.samples / 2;
    float[] samples = null;

    while (m_RecordingRoutine != 0 && m_Recording != null)
    {
      int writePos = Microphone.GetPosition(m_MicrophoneID);
      if (writePos > m_Recording.samples || !Microphone.IsRecording(m_MicrophoneID))
      {
        Log.Error("MicrophoneWidget", "Microphone disconnected.");

        StopRecording();
        yield break;
      }

      if ((bFirstBlock && writePos >= midPoint)
        || (!bFirstBlock && writePos < midPoint))
      {
        // front block is recorded, make a RecordClip and pass it onto our callback.
        samples = new float[midPoint];
        m_Recording.GetData(samples, bFirstBlock ? 0 : midPoint);

        AudioData record = new AudioData();
        record.MaxLevel = Mathf.Max(samples);
        record.Clip = AudioClip.Create("Recording", midPoint, m_Recording.channels, m_RecordingHZ, false);
        record.Clip.SetData(samples, 0);

        m_SpeechToText.OnListen(record);

        bFirstBlock = !bFirstBlock;
      }
      else
      {
        // calculate the number of samples remaining until we ready for a block of audio, 
        // and wait that amount of time it will take to record.
        int remaining = bFirstBlock ? (midPoint - writePos) : (m_Recording.samples - writePos);
        float timeRemaining = (float)remaining / (float)m_RecordingHZ;

        yield return new WaitForSeconds(timeRemaining);
      }

    }

    yield break;
  }

  private void OnRecognize(SpeechRecognitionEvent result)
  {
        Debug.Log("OnRecognize");
        if (result != null && result.results.Length > 0)
    {
      foreach (var res in result.results)
      {
        foreach (var alt in res.alternatives)
        {
          string text = alt.transcript;
                    Debug.Log(text);

          Log.Debug("ExampleStreaming", string.Format("{0} ({1}, {2:0.00})\n", text, res.final ? "Final" : "Interim", alt.confidence));
        }
      }
    }
  }
}

这是我添加以获取麦克风的行.我只是对其进行编辑，以在函数 RecordingHandler 中以零索引提供麦克风设备，该索引实际上为空(我不知道为什么，这是故意留下的还是错误).

and this is the line i add to get microphone. I just edit it to provide Microphone Device at zero index which was actually null (I don't know why, is this intentionally left or an error), in the function RecordingHandler .

 m_MicrophoneID = Microphone.devices[0];

但是不幸的是，它没有在Event OnRecognize 中显示任何输出日志，我认为它应该执行.

but unfortunately it is not showing any output log in EventOnRecognize which i think that it should execute.

几秒钟后(如我给定的音频长度5)，它将显示这些日志.我做错了，我无法理解如何发短信.

Wile it displaying these logs, after some seconds (as i given length 5 of the audio). What i am doing wrong, i am unable to understand that how speech to text.

[DEBUG] OnListenClosed(), State = DISCONNECTED 
[DEBUG] KeepAlive exited.

我还尝试了IBM Watson Speech To text Scene，它也没有显示任何内容.

I have also tried IBM Watson Speech To text Scene it is also not showing anything.

推荐答案

我还不能流式传输实时输出，但是可以通过watson服务将音频剪辑转换为文本，这是简单的代码(花了三点天).

I am not able to stream real-time output yet but become able to convert audio clip into text through watson service and here is the simple code (which took three days).

using UnityEngine;
using System.Collections;
using IBM.Watson.DeveloperCloud.Services.SpeechToText.v1;

public class AudioClipToTextWatson : MonoBehaviour {
    // Non-streaming
    SpeechToText m_SpeechToText = new SpeechToText();
    public AudioClip m_AudioClip = new AudioClip();
    public bool on = false;

    void Start () {
        m_AudioClip = Microphone.Start(Microphone.devices[0], false, 4, 44100);

            m_SpeechToText.Recognize(m_AudioClip, OnRecognize);
            //  Streaming
            m_SpeechToText.StartListening(OnRecognize);
            //  Stop listening
            m_SpeechToText.StopListening();
    }


    private void OnRecognize(SpeechRecognitionEvent result)
    {
        Debug.Log("result : " + result);
        if (result != null && result.results.Length > 0)
        {
            foreach (var res in result.results)
            {
                foreach (var alt in res.alternatives)
                {
                    string text = alt.transcript;
                    Debug.Log(text);
                    Debug.Log(res.final);
                }
            }
        }
    }

}

注意:您可以使用麦克风录制和音频片段，并将其转换为文本.如果您已经有音频，则将其放到检查器中，并在开始事件中的第一行中注释掉.

Note :You can record and an audio clip using your microphone and convert it to text. If you already have an audio then drop it to inspector and comment out the first line in Start Event.

这篇关于IBM Watson语音转文本服务未在Unity3d中给出响应的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！