Event listeners

Register a ResultListener to receive readiness, start, interim and final results, speech segment, input level, and error events for speech recognition.

Callback	Description
`onReady()`	Called when the SDK is ready to process a speech recognition request.
`onStartedRecognition()`	Called when speech recognition starts.
`onUpdated(stt, completed)`	Called when the recognized text is updated. If `completed` is `true`, the recognition result is final.
`onUpdatedEpdData(on, off)`	Called when EPD (End Point Detection) updates the speech start (`on`) and end (`off`) positions. The values are byte offsets in the input audio buffer.
`onUpdatedRms(rms)`	Called when the RMS (Root Mean Square) value of the input speech signal is updated. RMS represents signal magnitude and is not a dB value.
`onEndedRecognition()`	Called when speech recognition ends.
`onError(errorType)`	Called when an error occurs during processing.

Kotlin
Java

private val resultListener = object : ResultListener {
    override fun onReady() = Unit
    override fun onStartedRecognition() = Unit

    override fun onUpdated(stt: String, completed: Boolean) {
        // Handle an interim or final recognition result.
    }

    override fun onUpdatedEpdData(on: Long, off: Long) = Unit
    override fun onUpdatedRms(rms: Double) = Unit
    override fun onEndedRecognition() = Unit

    override fun onError(errorType: ErrorType) {
        // Handle the error.
    }
}

speechToText.addListener(resultListener)

private final ResultListener resultListener = new ResultListener() {
    @Override public void onReady() {}
    @Override public void onStartedRecognition() {}

    @Override
    public void onUpdated(String stt, boolean completed) {
        // Handle an interim or final recognition result.
    }

    @Override public void onUpdatedEpdData(long on, long off) {}
    @Override public void onUpdatedRms(double rms) {}
    @Override public void onEndedRecognition() {}

    @Override
    public void onError(ErrorType errorType) {
        // Handle the error.
    }
};

speechToText.addListener(resultListener);

Register the listener before calling request() or sendAudio() so that no recognition events are missed. When the listener is no longer needed or before releasing the SDK, remove the same instance that was registered.

speechToText.removeListener(resultListener)

For error handling guidance, see Error code.