LiveSpeechRecognizer печатает случайные слова

Question

Моя программа в основном по речи на текст. Я использую LiveSpeechRecognizer CMU Sphinx. Моя программа работает без ошибок, но не печатает слова. Я имею в виду, что он печатает случайные слова из моего файла грамматики.

Я использую мой Samsung мобильный наушники, как мой микрофон

Код:

/* 
* To change this license header, choose License Headers in Project Properties. 
* To change this template file, choose Tools | Templates 
* and open the template in the editor. 
*/ 

package SpeechRec; 
import java.io.IOException; 

import javax.sound.sampled.*; 

import edu.cmu.sphinx.api.*; 
/** 
* 
* @author divyanshu kunwar 
*/ 
public class SpeechRec{ 

    // Variables 
    private String result; 

    // Threads 
    Thread speechThread; 
    Thread resourcesThread; 

    // LiveRecognizer 
    private LiveSpeechRecognizer recognizer; 

    /** 
    * Constructor 
    */ 
    public SpeechRec() { 

     // Loading Message 

     // Configuration 
     Configuration configuration = new Configuration(); 

     // Load model from the jar 
     configuration.setAcousticModelPath("resource:/edu/cmu/sphinx/models/en-us/en-us"); 
     configuration.setDictionaryPath("resource:/edu/cmu/sphinx/models/en-us/cmudict-en-us.dict"); 

     // if you want to use LanguageModelPath disable the 3 lines after which 
     // are setting a custom grammar-> 



     // Grammar 
     configuration.setGrammarPath("resource:/grammars"); 
     configuration.setGrammarName("grammar"); 
     configuration.setUseGrammar(true); 



     try { 
      recognizer = new LiveSpeechRecognizer(configuration); 
     } catch (IOException ex) { 
     } 

     // Start recognition process pruning previously cached data. 
     recognizer.startRecognition(true); 

     // Start the Thread 
     startSpeechThread(); 
     startResourcesThread(); 
    } 

    /** 
    * Starting the main Thread of speech recognition 
    */ 
    protected void startSpeechThread() { 

     // alive? 
     if (speechThread != null && speechThread.isAlive()) 
      return; 

     // initialise 
     speechThread = new Thread(() -> { 
      try { 
       while (true) { 
        /* 
        * This method will return when the end of speech is 
        * reached. Note that the end pointer will determine the end 
        * of speech. 
        */ 
        SpeechResult speechResult = recognizer.getResult(); 
        if (speechResult != null) { 

         result = speechResult.getHypothesis(); 
         System.out.println("You said: [" + result + "]\n"); 
         // logger.log(Level.INFO, "You said: " + result + "\n") 

        } else{} 

       } 
      } catch (Exception ex) { 
      } 

      }); 

     // Start 
     speechThread.start(); 

    } 

    /** 
    * Starting a Thread that checks if the resources needed to the 
    * SpeechRecognition library are available 
    */ 
    protected void startResourcesThread() { 

     // alive? 
     if (resourcesThread != null && resourcesThread.isAlive()) 
      return; 

     resourcesThread = new Thread(() -> { 
      try { 

       // Detect if the microphone is available 
       while (true) { 
        if (AudioSystem.isLineSupported(Port.Info.MICROPHONE)) { 
         // logger.log(Level.INFO, "Microphone is available.\n") 
        } else { 
         // logger.log(Level.INFO, "Microphone is not 
         // available.\n") 

        } 

        // Sleep some period 
        Thread.sleep(3500); 
       } 

      } catch (InterruptedException ex) { 
       resourcesThread.interrupt(); 
      } 
     }); 

     // Start 
     resourcesThread.start(); 
    } 

    /** 
    * Takes a decision based on the given result 
    */ 
    public void makeDesicion(String result) { 
     //implemented in the part 2 
    } 

    /** 
    * Java Main Application Method 
    * 
    * @param args 
    */ 
    public static void main(String[] args) { 

     // // Be sure that the user can't start this application by not giving 
     // the 
     // // correct entry string 
     // if (args.length == 1 && "SPEECH".equalsIgnoreCase(args[0])) 
     new SpeechRec(); 
     // else 
     // Logger.getLogger(Main.class.getName()).log(Level.WARNING, "Give me 
     // the correct entry string.."); 

    } 

} 

Grammar File: 
#JSGF V1.0; 
grammar grammar; 

public <numbers> = (one | two | three| four| five | six | seven | eight | nine | ten); 
public <words>=(how | where); 


Output: 
You said: [one] 

21:01:38.251 INFO liveCMN    35.23 -8.72 -11.38 11.38 2.68 -13.93 3.08 1.74 -8.43 -1.72 -0.49 -8.30 -3.83 
21:01:38.519 INFO speedTracker   This Time Audio: 0.76s Proc: 0.73s Speed: 0.96 X real time 
21:01:38.519 INFO speedTracker   Total Time Audio: 2.93s Proc: 0.99s 0.34 X real time 

You said: [one] 

21:01:38.519 INFO memoryTracker   Mem Total: 434.00 Mb Free: 303.12 Mb 
21:01:38.519 INFO memoryTracker   Used: This: 130.88 Mb Avg: 124.48 Mb Max: 130.88 Mb 
21:01:40.233 INFO speedTracker   This Time Audio: 0.30s Proc: 0.19s Speed: 0.64 X real time 
You said: [two] 

21:01:40.233 INFO speedTracker   Total Time Audio: 3.23s Proc: 1.18s 0.37 X real time 
21:01:40.233 INFO memoryTracker   Mem Total: 434.00 Mb Free: 295.40 Mb 
21:01:40.233 INFO memoryTracker   Used: This: 138.60 Mb Avg: 129.18 Mb Max: 138.60 Mb 
21:01:41.777 INFO liveCMN    35.55 -8.33 -10.84 13.40 3.05 -13.95 4.24 1.77 -9.87 -2.61 -1.27 -8.00 -3.56 
21:01:42.853 INFO liveCMN    35.35 -7.67 -10.27 11.74 2.97 -13.25 3.25 1.60 -9.12 -2.91 -1.70 -7.04 -3.10 
21:01:42.990 INFO speedTracker   This Time Audio: 1.29s Proc: 1.25s Speed: 0.97 X real time 
21:01:42.990 INFO speedTracker   Total Time Audio: 4.52s Proc: 2.43s 0.54 X real time 

You said: [one] 

21:01:42.990 INFO memoryTracker   Mem Total: 434.00 Mb Free: 279.92 Mb 
21:01:42.991 INFO memoryTracker   Used: This: 154.08 Mb Avg: 135.41 Mb Max: 154.08 Mb 

Я не сказал эти слова. Пожалуйста помоги.

Answer 1

Получите правильный микрофон, который записывается без шума.