I am working on app which require speech to text. I have integrated SpeechRecognizer service. Please check below demo project code for same. Here I have tested that SpeechRecognizer stop listening automatically after 10 to 15 seconds. So I looked for following solution to increase the listening time but did not worked.
Solution 1: SpeechRecognizer - time limit
- For the above link i found the problem is text getting cut like Said "Hello , what is the serial number of your currency, It is ABC5654548585" so in first result i got Hello, What is the serial number of your currency and in second i got it is ABC5654548585".So I am unable to parse response properly.
Solution 2: I set the following parameters but did not worked.
- EXTRA_SPEECH_INPUT_COMPLETE_SILENCE_LENGTH_MILLIS EXTRA_SPEECH_INPUT_MINIMUM_LENGTH_MILLIS EXTRA_SPEECH_INPUT_POSSIBLY_COMPLETE_SILENCE_LENGTH_MILLIS.
Solution 3: I also integrated mozilla speech to text but it is not accurate as google.
Please help me I am stucked here.Not able to proceed any more.
class MainActivity : AppCompatActivity() {
lateinit var speechRecognize: SpeechRecognizer
var editText: EditText? = null
var micButton: ImageView? = null
var speechRecognizerIntent: Intent? = null
var audioManager: AudioManager? = null
var textRecorded = ""
var isActive = true
override fun onCreate(savedInstanceState: Bundle?) {
super.onCreate(savedInstanceState)
setContentView(R.layout.activity_main)
speechRecognize = SpeechRecognizer.createSpeechRecognizer(this)
audioManager = getSystemService(Context.AUDIO_SERVICE) as AudioManager?
muteRecognition(true)
editText = findViewById(R.id.text);
micButton = findViewById(R.id.button);
micButton?.setOnClickListener {
if (it.tag == null) {
isActive = true
launchSpeechIntent()
speechRecognize.startListening(speechRecognizerIntent)
it.tag = 1
} else if (it.tag == 1) {
isActive = false
parseText()
speechRecognize.stopListening()
speechRecognize.destroy()
it.tag = null
}
}
val permissions = ArrayList<String>()
if (ActivityCompat.checkSelfPermission(
this,
Manifest.permission.RECORD_AUDIO
)
!= PackageManager.PERMISSION_GRANTED
) {
permissions.add(Manifest.permission.RECORD_AUDIO)
}
if (ActivityCompat.checkSelfPermission(
this,
Manifest.permission.WRITE_EXTERNAL_STORAGE
)
!= PackageManager.PERMISSION_GRANTED
) {
permissions.add(Manifest.permission.WRITE_EXTERNAL_STORAGE)
}
if (permissions.isNotEmpty()) {
ActivityCompat.requestPermissions(
this,
permissions.toArray(arrayOf<String>()),
10
)
} else
launchSpeechIntent()
//else
// startRequest()
}
@Suppress("DEPRECATION")
private fun muteRecognition(mute: Boolean) {
audioManager?.let {
if (Build.VERSION.SDK_INT >= Build.VERSION_CODES.M) {
val flag = if (mute) AudioManager.ADJUST_MUTE else AudioManager.ADJUST_UNMUTE
it.adjustStreamVolume(AudioManager.STREAM_NOTIFICATION, flag, 0)
it.adjustStreamVolume(AudioManager.STREAM_ALARM, flag, 0)
it.adjustStreamVolume(AudioManager.STREAM_MUSIC, flag, 0)
it.adjustStreamVolume(AudioManager.STREAM_RING, flag, 0)
it.adjustStreamVolume(AudioManager.STREAM_SYSTEM, flag, 0)
} else {
it.setStreamMute(AudioManager.STREAM_NOTIFICATION, mute)
it.setStreamMute(AudioManager.STREAM_ALARM, mute)
it.setStreamMute(AudioManager.STREAM_MUSIC, mute)
it.setStreamMute(AudioManager.STREAM_RING, mute)
it.setStreamMute(AudioManager.STREAM_SYSTEM, mute)
}
}
}
fun launchSpeechIntent() {
speechRecognizerIntent = Intent(RecognizerIntent.ACTION_RECOGNIZE_SPEECH);
speechRecognizerIntent?.putExtra(
RecognizerIntent.EXTRA_LANGUAGE_MODEL,
RecognizerIntent.LANGUAGE_MODEL_FREE_FORM
);
speechRecognizerIntent?.putExtra(RecognizerIntent.EXTRA_CALLING_PACKAGE, this.packageName)
speechRecognizerIntent?.putExtra(RecognizerIntent.EXTRA_LANGUAGE, Locale.getDefault());
speechRecognizerIntent?.putExtra(RecognizerIntent.EXTRA_MAX_RESULTS, 1)
speechRecognizerIntent?.putExtra(
RecognizerIntent.EXTRA_SPEECH_INPUT_POSSIBLY_COMPLETE_SILENCE_LENGTH_MILLIS,
30000
)
speechRecognize.setRecognitionListener(object : RecognitionListener {
override fun onReadyForSpeech(params: Bundle?) {
Log.e("ready for speeach", "true")
}
override fun onRmsChanged(rmsdB: Float) {
Log.e("RMS changed", rmsdB.toString())
}
override fun onBufferReceived(buffer: ByteArray?) {
Log.e("buffer", buffer.toString())
}
override fun onPartialResults(partialResults: Bundle?) {
Log.e("ready for speeach", "true" + partialResults.toString())
}
override fun onEvent(eventType: Int, params: Bundle?) {
Log.e("event", eventType.toString())
Log.e("params", params.toString())
}
override fun onBeginningOfSpeech() {
editText!!.setHint("Listening...........")
editText!!.setText("")
}
override fun onEndOfSpeech() {
}
override fun onError(error: Int) {
Log.e("Error", error.toString())
speechRecognize.startListening(speechRecognizerIntent)
//editText?.setText("Error:"+error.toString())
}
override fun onResults(results: Bundle?) {
val data: ArrayList<String>? =
results!!.getStringArrayList(SpeechRecognizer.RESULTS_RECOGNITION)
var flot = results!!.getFloatArray(SpeechRecognizer.CONFIDENCE_SCORES)
textRecorded += "#${data!!.first()}"
// restart()
}
})
}
fun restart() {
if (isActive) {
speechRecognize.destroy()
launchSpeechIntent()
speechRecognize.startListening(speechRecognizerIntent)
} else {
speechRecognize.stopListening()
speechRecognize.destroy()
}
}
override fun onResume() {
super.onResume()
}
fun parseText() {
try {
editText?.setText(textRecorded)
} catch (e: Exception) {
e.printStackTrace()
}
}
private fun checkPermission() {
if (Build.VERSION.SDK_INT >= Build.VERSION_CODES.M) {
ActivityCompat.requestPermissions(
this,
arrayOf(Manifest.permission.RECORD_AUDIO),
10
)
}
}
override fun onRequestPermissionsResult(
requestCode: Int,
permissions: Array<out String>,
grantResults: IntArray
) {
super.onRequestPermissionsResult(requestCode, permissions, grantResults)
when (requestCode) {
10 ->
if (grantResults[0] != PackageManager.PERMISSION_GRANTED) {
launchSpeechIntent()
}
}
}
override fun onDestroy() {
super.onDestroy()
speechRecognize.stopListening()
speechRecognize.destroy()
}
}
