CMU Sphinx / Forums / Help: Android demo app not working

callintegrator - 2014-06-13

tried to get the Android demo application up and running. onBeginningOfSpeech and onEndOfSpeech are being called regularly.But never onResult or onPartialResult. As a result I am stuck on the screen where I have to say "oh mighty computer". What might be the reason for this?

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
- Nickolay V. Shmyrev - 2014-06-13
  
  Demo creates raw files in on sdcard in the folder pointed by setRawLogDir in sources. Share them.
  
  If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

callintegrator - 2014-06-13

here we go

Last edit: callintegrator 2014-06-13

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
- Nickolay V. Shmyrev - 2014-06-13
  
  Demo waits for "oh mighty computer" to activate decoding.
  
  I don't hear you say "oh mighty computer" in the raw file. Start the application, say "oh mighty computer" multiple times to activate the decoding. Share the raw files again.
  
  If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

callintegrator - 2014-06-13

I said it quite a few times, I did it again and again. Now I got it convinced to recognize, at least some times. The recognition afterwards is as bad as the recognition of "oh mighty computer". Can the system be trained on Android? I only need it to recognize a few phrases, but that reliably. I am fine to train it to recognize my accent.

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

callintegrator - 2014-06-13

PS: Can I analyze the raw file myself to see what is being said?

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Nickolay V. Shmyrev - 2014-06-14

The recognition afterwards is as bad as the recognition of "oh mighty computer".

You need to share raw files in order to get help on this issue

Can the system be trained on Android? I only need it to recognize a few phrases, but that reliably. I am fine to train it to recognize my accent.

Yes, things could be significantly improved

PS: Can I analyze the raw file myself to see what is being said?

Yes, they are simply raw audio files. You can open them in Audacity or in Wavesurfer.

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
- callintegrator - 2014-06-14
  
  Can the system be trained on Android? I only need it to recognize a few phrases, but >>that reliably. I am fine to train it to recognize my accent.
  
  Yes, things could be significantly improved
  
  How can I do this?
  
  If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
  - Nickolay V. Shmyrev - 2014-06-14
    
    How can I do this?
    
    You need to share raw files
    
    If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

callintegrator - 2014-06-16

https://anonfiles.com/file/74530c8ff2fa278aa5a80df9f5dd7a38

In these files you should find about 10 times "oh might computer", once it recognized it and I was able to say "digits" and 1,2, then it switched back to "oh mighty computer".

but how shall that help on terms I need like "item 1" or "paint" or whatever? Especially in an app you have different users, so each user should be able to say these words once at startup for a better recognition.

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
- Nickolay V. Shmyrev - 2014-06-17
  
  Well, we can't guarantee you a good accuracy for arbitrary accent yet. To improve "oh mighty computer" recognition you need to make the following changes in the dictionary:
  
  oh AO UW computer K AO M P Y UW T AO R
  
  Also make sure that kwsThreshold is configured in sources in recognizer setup method to 1e-40, not 1e-20.
  
  If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Jeyamaal - 2017-06-01

I have downloaded the demo application for Android. from https://github.com/cmusphinx/pocketsphinx-android-demo.

I'm also encoutered the same problems which are mentioned before.
https://sourceforge.net/p/cmusphinx/discussion/help/thread/3644b282/#34dc

https://sourceforge.net/p/cmusphinx/discussion/help/thread/3644b282/#2b49

And I want to recognize my own words also (Continous Speech).

https://sourceforge.net/p/cmusphinx/discussion/help/thread/3644b282/#ec5f

Could anyone help me??

Last edit: Jeyamaal 2017-06-01

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
- Nickolay V. Shmyrev - 2017-06-01
  
  Sure, as soon as you describe the problem you have in more details.
  
  If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
  - Jeyamaal - 2017-06-01
    
    1.I said "Oh mighty computer" several times, I did it again and again.
    But it coudn't recogniz the words.
    2.Raw files is not created in the destinated folder.
    
    Please check my Log details.
    
    6-01 20:23:55.571 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: dict.c(213): Dictionary size 133425, allocated 0 KiB for strings, 0 KiB for phones
    06-01 20:23:55.571 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: dict.c(361): 5 words read
    06-01 20:23:55.571 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: dict2pid.c(396): Building PID tables for dictionary
    06-01 20:23:55.571 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: dict2pid.c(406): Allocating 42^3 * 2 bytes (144 KiB) for word-initial triphones
    06-01 20:23:55.621 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: dict2pid.c(132): Allocated 21336 bytes (20 KiB) for word-final triphones
    06-01 20:23:55.631 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: dict2pid.c(196): Allocated 21336 bytes (20 KiB) for single-phone word triphones
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: kws_search.c(406): KWS(beam: -1080, plp: -23, default threshold -450, delay 10)
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/SpeechRecognizer: Load JSGF /storage/emulated/0/Android/data/edu.cmu.sphinx.pocketsphinx/files/sync/menu.gram
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: jsgf.c(706): Defined rule: PUBLIC <menu.item>
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_model.c(208): Computing transitive closure for null transitions
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_model.c(270): 0 null transitions added
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_search.c(227): FSG(beam: -1080, pbeam: -1080, wbeam: -634; wip: -26, pip: 0)
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_model.c(423): Adding silence transitions for <sil> to FSG
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_model.c(443): Added 2 silence word transitions
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_model.c(423): Adding silence transitions for <sil> to FSG
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_model.c(443): Added 2 silence word transitions
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_model.c(423): Adding silence transitions for [NOISE] to FSG
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_model.c(443): Added 2 silence word transitions
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_search.c(173): Added 1 alternate word transitions
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_lextree.c(110): Allocated 172 bytes (0 KiB) for left and right context phones
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_lextree.c(256): 25 HMM nodes in lextree (8 leaves)
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_lextree.c(259): Allocated 3000 bytes (2 KiB) for all lextree nodes
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_lextree.c(262): Allocated 960 bytes (0 KiB) for lextree leafnodes
    06-01 20:23:55.681 12924-12975/edu.cmu.sphinx.pocketsphinx I/SpeechRecognizer: Load JSGF /storage/emulated/0/Android/data/edu.cmu.sphinx.pocketsphinx/files/sync/digits.gram
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: jsgf.c(706): Defined rule: <digits.digit>
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: jsgf.c(706): Defined rule: <digits.g00001>
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: jsgf.c(706): Defined rule: PUBLIC <digits.digits>
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: jsgf.c(365): Right recursion <digits.g00001> 2 => 0
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_model.c(208): Computing transitive closure for null transitions
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_model.c(270): 0 null transitions added
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_search.c(227): FSG(beam: -1080, pbeam: -1080, wbeam: -634; wip: -26, pip: 0)
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_model.c(423): Adding silence transitions for <sil> to FSG
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_model.c(443): Added 3 silence word transitions
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_model.c(423): Adding silence transitions for <sil> to FSG
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_model.c(443): Added 3 silence word transitions
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_model.c(423): Adding silence transitions for [NOISE] to FSG
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_model.c(443): Added 3 silence word transitions
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_search.c(173): Added 4 alternate word transitions
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_lextree.c(110): Allocated 258 bytes (0 KiB) for left and right context phones
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_lextree.c(256): 191 HMM nodes in lextree (163 leaves)
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_lextree.c(259): Allocated 22920 bytes (22 KiB) for all lextree nodes
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: fsg_lextree.c(262): Allocated 19560 bytes (19 KiB) for lextree leafnodes
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/SpeechRecognizer: Load N-gram model /storage/emulated/0/Android/data/edu.cmu.sphinx.pocketsphinx/files/sync/weather.dmp
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_model_trie.c(354): Trying to read LM in trie binary format
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_model_trie.c(365): Header doesn't match
    06-01 20:23:55.691 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_model_trie.c(177): Trying to read LM in arpa format
    06-01 20:23:55.701 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_model_trie.c(70): No \data\ mark in LM file
    06-01 20:23:55.701 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_model_trie.c(445): Trying to read LM in dmp format
    06-01 20:23:55.701 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_model_trie.c(527): ngrams 1=779, 2=14416, 3=46976
    06-01 20:23:56.022 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: lm_trie.c(474): Training quantizer
    06-01 20:23:56.202 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: lm_trie.c(482): Building LM trie
    06-01 20:24:01.618 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_search_fwdtree.c(74): Initializing search tree
    06-01 20:24:01.628 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_search_fwdtree.c(101): 788 unique initial diphones
    06-01 20:24:01.628 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_search_fwdtree.c(186): Creating search channels
    06-01 20:24:01.668 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_search_fwdtree.c(323): Max nonroot chan increased to 2220
    06-01 20:24:01.668 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_search_fwdtree.c(333): Created 256 root, 2092 non-root channels, 9 single-phone words
    06-01 20:24:01.668 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_search_fwdflat.c(157): fwdflat: min_ef_width = 4, max_sf_win = 25
    06-01 20:24:01.668 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_model_trie.c(354): Trying to read LM in trie binary format
    06-01 20:24:01.668 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_model_trie.c(365): Header doesn't match
    06-01 20:24:01.668 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_model_trie.c(177): Trying to read LM in arpa format
    06-01 20:24:01.668 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_model_trie.c(70): No \data\ mark in LM file
    06-01 20:24:01.668 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_model_trie.c(445): Trying to read LM in dmp format
    06-01 20:24:01.668 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: ngram_model_trie.c(527): ngrams 1=43, 2=1509, 3=21837
    06-01 20:24:01.708 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: lm_trie.c(474): Training quantizer
    06-01 20:24:01.748 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: lm_trie.c(482): Building LM trie
    06-01 20:24:01.778 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: allphone_search.c(236): Building PHMM net of 42 phones
    06-01 20:24:01.778 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: allphone_search.c(309): 42 nodes, 1764 links
    06-01 20:24:01.778 12924-12975/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: allphone_search.c(603): Allphone(beam: -1080, pbeam: -1080)
    06-01 20:24:01.778 12924-12924/edu.cmu.sphinx.pocketsphinx I/SpeechRecognizer: Start recognition "wakeup"
    06-01 20:24:01.888 12924-13478/edu.cmu.sphinx.pocketsphinx D/SpeechRecognizer: Starting decoding
    06-01 20:24:01.888 12924-13478/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: pocketsphinx.c(986): Writing raw audio file: /storage/emulated/0/Android/data/edu.cmu.sphinx.pocketsphinx/files/sync/000000000.raw
    06-01 20:24:15.492 12924-13478/edu.cmu.sphinx.pocketsphinx I/cmusphinx: INFO: cmn_live.c(88): Update from <
    06-01 20:24:15.492 12924-13478/edu.cmu.sphinx.pocketsphinx I/cmusphinx: 40.00
    06-01 20:24:15.492 12924-13478/edu.cmu.sphinx.pocketsphinx I/cmusphinx: 10.00
    06-01 20:24:15.492 12924-13478/edu.cmu.sphinx.pocketsphinx I/cmusphinx: 10.00
    06-01 20:24:15.492 12924-13478/edu.cmu.sphinx.pocketsphinx I/cmusphinx: 0.00
    06-01 20:24:15.492 12924-13478/edu.cmu.sphinx.pocketsphinx I/cmusphinx: 0.00
    06-01 20:24:15.492 12924-13478/edu.cmu.sphinx.pocketsphinx I/cmusphinx: 0.00
    06-01 20:24:15.492 12924-13478/edu.cmu.sphinx.pocketsphinx I/cmusphinx: 0.00
    06-01 20:24:15.492 12924-13478/edu.cmu.sphinx.pocketsphinx I/cmusphinx: 0.00
    06-01 20:24:15.492 12924-13478/edu.cmu.sphinx.pocketsphinx I/cmusphinx: 0.00
    06-01 20:24:15.492 12924-13478/edu.cmu.sphinx.pocketsphinx I/cmusphinx: 0.00
    06-01 20:24:15.492 12924-13478/edu.cmu.sphinx.pocketsphinx I/cmusphinx: 0.00
    06-01 20:24:15.492 12924-13478/edu.cmu.sphinx.pocketsphinx I/cmusphinx: 0.00
    06-01 20:24:15.492 12924-13478/edu.cmu.sphinx.pocketsphinx I/cmusphinx: 0.00
    * 06-01 20:24:15.492 12924-13478/edu.cmu.sphinx.pocketsphinx I/cmusphinx: >
    
    If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
    - Nickolay V. Shmyrev - 2017-06-01
      
      Raw files must be there, your log says
      
      Writing raw audio file: /storage/emulated/0/Android/data/edu.cmu.sphinx.pocketsphinx/files/sync/000000000.raw
      
      You need to share them. Also you need to try to run the system for longer time, not just for 5 seconds
      
      If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
      - Jeyamaal - 2017-06-01
        
        Some times its recognize , but most of the time it couldn't recognized.
        
        "You need to share them"
        I attach here with my raw file.
        
        000000009.raw
        
        If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
        
        Nickolay V. Shmyrev - 2017-06-02
        
        You say simply "computer", it expects the keyphrase "oh mighty computer" without pauses. And pronunciation should be closer to US English.
        
        If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

I wrote to identify the "hello" keyword to initiate the voice to text, then it should convert voice to text continuoulsy.But it only identify the "hello".

My purpose is to continoulsy do the voice to text process.

Could you please help me??

Please check my code below.

package edu.cmu.pocketsphinx.demo;

import android.Manifest;
import android.app.Activity;
import android.content.pm.PackageManager;
import android.os.AsyncTask;
import android.os.Bundle;
import android.support.v4.app.ActivityCompat;
import android.support.v4.content.ContextCompat;
import android.util.Log;
import android.widget.TextView;
import android.widget.Toast;

import java.io.File;
import java.io.IOException;
import java.util.HashMap;

import edu.cmu.pocketsphinx.Assets;
import edu.cmu.pocketsphinx.Hypothesis;
import edu.cmu.pocketsphinx.RecognitionListener;
import edu.cmu.pocketsphinx.SpeechRecognizer;
import edu.cmu.pocketsphinx.SpeechRecognizerSetup;

import static android.widget.Toast.makeText;

public class MainActivity extends Activity implements
RecognitionListener {

/* Named searches allow to quickly reconfigure the decoder */
private static final String KWS_SEARCH = "wakeup";

private static final String KEYPHRASE = "hello";

private static final String FORECAST_SEARCH = "forecast";
private static final String DIGITS_SEARCH = "digits";
private static final String PHONE_SEARCH = "phones";
private static final String MENU_SEARCH = "menu";

/* Used to handle permission request */
private static final int PERMISSIONS_REQUEST_RECORD_AUDIO = 1;

private SpeechRecognizer recognizer;
private HashMap<String, Integer> captions;

private File appDir;

@Override
public void onCreate(Bundle state) {
    super.onCreate(state);
    setContentView(R.layout.main);

    try {
        Log.d("Tag","before trying to sync assets");
        Assets assets = new Assets(MainActivity.this);
        appDir = assets.syncAssets();
    } catch (IOException e) {
        throw new RuntimeException("failed to synchronize assets", e);
    }

    try {

        Log.d("TAG", "before recognizer instantiaiton");
        recognizer = SpeechRecognizerSetup.defaultSetup()
                .setAcousticModel(new File(appDir, "en-us-ptm"))
                .setDictionary(new File(appDir, "cmudict-en-us.dict"))
                .setRawLogDir(appDir) // To disable logging of raw audio comment out this call (takes a lot of space on the device)
                .setKeywordThreshold(1e-40f)
                .getRecognizer();

        recognizer.addListener(this);
        recognizer.addKeyphraseSearch(KWS_SEARCH, KEYPHRASE);
        recognizer.startListening(KWS_SEARCH);

    }catch(Exception e){

        e.printStackTrace();
    }
}

@Override
public void onPartialResult(Hypothesis hyp) {

    if(hyp==null)
        return;

    //Restart the recognition if keyword is found

    String text=hyp.getHypstr();
    Log.d("Spoken text",text);
    ((TextView) findViewById(R.id.result_text)).setText(text);
    recognizer.cancel();
    recognizer.startListening(KWS_SEARCH);

}

@Override
public void onDestroy() {
    super.onDestroy();

    if (recognizer != null) {
        recognizer.cancel();
        recognizer.shutdown();
    }
}

@Override
public void onResult(Hypothesis hypothesis) {
    ((TextView) findViewById(R.id.result_text)).setText("");
    if (hypothesis != null) {
        String text = hypothesis.getHypstr();
        makeText(getApplicationContext(), text, Toast.LENGTH_SHORT).show();
    }
}

@Override
public void onBeginningOfSpeech() {
}

/**
 * We stop recognizer here to get a final result
 */
@Override
public void onEndOfSpeech() {
    if (!recognizer.getSearchName().equals(KWS_SEARCH))
        switchSearch(KWS_SEARCH);
}

private void switchSearch(String searchName) {
    recognizer.stop();

    // If we are not spotting, start listening with timeout (10000 ms or 10 seconds).
    if (searchName.equals(KWS_SEARCH))
        recognizer.startListening(searchName);
    else
        recognizer.startListening(searchName, 10000);

    String caption = getResources().getString(captions.get(searchName));
    ((TextView) findViewById(R.id.caption_text)).setText(caption);
}

@Override
public void onError(Exception error) {
    ((TextView) findViewById(R.id.caption_text)).setText(error.getMessage());
}

@Override
public void onTimeout() {
    switchSearch(KWS_SEARCH);
}

}

Nickolay V. Shmyrev - 2017-06-03

It is not possible

https://stackoverflow.com/questions/25949295/cmusphinx-pocketsphinx-recognize-all-or-large-amount-of-words

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Maaz Patel - 2017-11-22

Hello Sir, You have done a great job.
I havce tried the demo its working fine . but I am having some doubts it would helpful for me if you resolved it.
1) when I tried my key pharese "help", it will detect the help word without saying a single word. though I have made my cabin noise frees still it detects the help word without saying anything. I don't undestand how to resolve this.
2) Do I need to add someting in dictionary to reognize my word "help".
3) Whats is the realtionship between keypharse and keyword.

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
- Nickolay V. Shmyrev - 2017-11-22
  
  1) when I tried my key pharese "help", it will detect the help word without saying a single word. though I have made my cabin noise frees still it detects the help word without saying anything. I don't undestand how to resolve this.
  
  This keyphrase is too short for reliable detection. Tutorial recommends to use keyword of 3-5 syllables. You need to read the tutorial first.
  
  http://cmusphinx.github.io/wiki/tutoriallm
  
  3) Whats is the realtionship between keypharse and keyword.
  
  keyphase consist of several words like "oh mighty computer".
  
  If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Maaz Patel - 2017-11-23

Hi,
thanks for the soloution.
I am playing the music in the background and at the same time if I say "Oh Mighty Computer" it dosent recognize it. what should I do in that case? My requirement is such that it will play music continuesly in background and meanwhile if I say "Oh Mighty Computer" it should detect it. how do I remove noise from the system. the music is played in the same device through which we are listening.

Can you have solution for this?

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
- Nickolay V. Shmyrev - 2017-11-23
  
  Pocketsphinx does not support such thing right now. That would be a problem with a PhD disseration.
  
  If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Android demo app not working

Speech Recognition Toolkit

Forums

Help

Android demo app not working document.SUBSCRIPTION_OPTIONS = { "thing": "topic", "subscribed": false, "url": "subscribe", "icon": { "css": "fa fa-envelope-o" } };

Android demo app not working