Examples with RecognizeWithWebsocketsOptions com.ibm.watson.speech_to_text.v1.model.RecognizeWithWebsocketsOptions used on opensource projects

Example 1 with RecognizeWithWebsocketsOptions

use of com.ibm.watson.speech_to_text.v1.model.RecognizeWithWebsocketsOptions in project openhab-addons by openhab.

the class WatsonSTTService method recognize.

@Override
public STTServiceHandle recognize(STTListener sttListener, AudioStream audioStream, Locale locale, Set<String> set) throws STTException {
    if (config.apiKey.isBlank() || config.instanceUrl.isBlank()) {
        throw new STTException("service is not correctly configured");
    }
    String contentType = getContentType(audioStream);
    if (contentType == null) {
        throw new STTException("Unsupported format, unable to resolve audio content type");
    }
    logger.debug("Content-Type: {}", contentType);
    var speechToText = new SpeechToText(new IamAuthenticator.Builder().apikey(config.apiKey).build());
    speechToText.setServiceUrl(config.instanceUrl);
    if (config.optOutLogging) {
        speechToText.setDefaultHeaders(Map.of("X-Watson-Learning-Opt-Out", "1"));
    }
    RecognizeWithWebsocketsOptions wsOptions = new RecognizeWithWebsocketsOptions.Builder().audio(audioStream).contentType(contentType).redaction(config.redaction).smartFormatting(config.smartFormatting).model(locale.toLanguageTag() + "_BroadbandModel").interimResults(true).backgroundAudioSuppression(config.backgroundAudioSuppression).speechDetectorSensitivity(config.speechDetectorSensitivity).inactivityTimeout(config.inactivityTimeout).build();
    final AtomicReference<@Nullable WebSocket> socketRef = new AtomicReference<>();
    final AtomicBoolean aborted = new AtomicBoolean(false);
    executor.submit(() -> {
        int retries = 2;
        while (retries > 0) {
            try {
                socketRef.set(speechToText.recognizeUsingWebSocket(wsOptions, new TranscriptionListener(sttListener, config, aborted)));
                break;
            } catch (RuntimeException e) {
                var cause = e.getCause();
                if (cause instanceof SSLPeerUnverifiedException) {
                    logger.debug("Retrying on error: {}", cause.getMessage());
                    retries--;
                } else {
                    var errorMessage = e.getMessage();
                    logger.warn("Aborting on error: {}", errorMessage);
                    sttListener.sttEventReceived(new SpeechRecognitionErrorEvent(errorMessage != null ? errorMessage : "Unknown error"));
                    break;
                }
            }
        }
    });
    return new STTServiceHandle() {

        @Override
        public void abort() {
            if (!aborted.getAndSet(true)) {
                var socket = socketRef.get();
                if (socket != null) {
                    socket.close(1000, null);
                    socket.cancel();
                    try {
                        Thread.sleep(100);
                    } catch (InterruptedException ignored) {
                    }
                }
            }
        }
    };
}

Example 2 with RecognizeWithWebsocketsOptions

use of com.ibm.watson.speech_to_text.v1.model.RecognizeWithWebsocketsOptions in project java-sdk by watson-developer-cloud.

the class SpeechToText method recognizeUsingWebSocket.

/**
 * Sends audio and returns transcription results for recognition requests over a WebSocket
 * connection. Requests and responses are enabled over a single TCP connection that abstracts much
 * of the complexity of the request to offer efficient implementation, low latency, high
 * throughput, and an asynchronous response. By default, only final results are returned for any
 * request; to enable interim results, set the interimResults parameter to true.
 *
 * <p>The service imposes a data size limit of 100 MB per utterance (per recognition request). You
 * can send multiple utterances over a single WebSocket connection. The service automatically
 * detects the endianness of the incoming audio and, for audio that includes multiple channels,
 * downmixes the audio to one-channel mono during transcoding. (For the audio/l16 format, you can
 * specify the endianness.)
 *
 * @param recognizeOptions the recognize options
 * @param callback the {@link RecognizeCallback} instance where results will be sent
 * @return the {@link WebSocket}
 */
public WebSocket recognizeUsingWebSocket(RecognizeWithWebsocketsOptions recognizeOptions, RecognizeCallback callback) {
    com.ibm.cloud.sdk.core.util.Validator.notNull(recognizeOptions, "recognizeOptions cannot be null");
    com.ibm.cloud.sdk.core.util.Validator.notNull(recognizeOptions.audio(), "audio cannot be null");
    com.ibm.cloud.sdk.core.util.Validator.notNull(callback, "callback cannot be null");
    HttpUrl.Builder urlBuilder = HttpUrl.parse(getServiceUrl() + "/v1/recognize").newBuilder();
    if (recognizeOptions.model() != null) {
        urlBuilder.addQueryParameter("model", recognizeOptions.model());
    }
    if (recognizeOptions.customizationId() != null) {
        urlBuilder.addQueryParameter("customization_id", recognizeOptions.customizationId());
    }
    if (recognizeOptions.languageCustomizationId() != null) {
        urlBuilder.addQueryParameter("language_customization_id", recognizeOptions.languageCustomizationId());
    }
    if (recognizeOptions.acousticCustomizationId() != null) {
        urlBuilder.addQueryParameter("acoustic_customization_id", recognizeOptions.acousticCustomizationId());
    }
    if (recognizeOptions.baseModelVersion() != null) {
        urlBuilder.addQueryParameter("base_model_version", recognizeOptions.baseModelVersion());
    }
    String url = urlBuilder.toString().replace("https://", "wss://");
    Request.Builder builder = new Request.Builder().url(url);
    setAuthentication(builder);
    setDefaultHeaders(builder);
    OkHttpClient client = configureHttpClient();
    return client.newWebSocket(builder.build(), new SpeechToTextWebSocketListener(recognizeOptions, callback));
}

Example 3 with RecognizeWithWebsocketsOptions

use of com.ibm.watson.speech_to_text.v1.model.RecognizeWithWebsocketsOptions in project java-sdk by watson-developer-cloud.

the class RecognizeUsingWebSocketsExample method main.

public static void main(String[] args) throws FileNotFoundException, InterruptedException {
    Authenticator authenticator = new IamAuthenticator("<iam_api_key>");
    SpeechToText service = new SpeechToText(authenticator);
    FileInputStream audio = new FileInputStream("src/test/resources/speech_to_text/sample1.wav");
    RecognizeWithWebsocketsOptions options = new RecognizeWithWebsocketsOptions.Builder().audio(audio).interimResults(true).contentType(HttpMediaType.AUDIO_WAV).build();
    service.recognizeUsingWebSocket(options, new BaseRecognizeCallback() {

        @Override
        public void onTranscription(SpeechRecognitionResults speechResults) {
            System.out.println(speechResults);
        }

        @Override
        public void onDisconnected() {
            lock.countDown();
        }
    });
    lock.await(1, TimeUnit.MINUTES);
}

Example 4 with RecognizeWithWebsocketsOptions

use of com.ibm.watson.speech_to_text.v1.model.RecognizeWithWebsocketsOptions in project java-sdk by watson-developer-cloud.

the class MicrophoneWithWebSocketsExample method main.

/**
 * The main method.
 *
 * @param args the arguments
 * @throws Exception the exception
 */
public static void main(final String[] args) throws Exception {
    Authenticator authenticator = new IamAuthenticator("<iam_api_key>");
    SpeechToText service = new SpeechToText(authenticator);
    // Signed PCM AudioFormat with 16kHz, 16 bit sample size, mono
    int sampleRate = 16000;
    AudioFormat format = new AudioFormat(sampleRate, 16, 1, true, false);
    DataLine.Info info = new DataLine.Info(TargetDataLine.class, format);
    if (!AudioSystem.isLineSupported(info)) {
        System.out.println("Line not supported");
        System.exit(0);
    }
    TargetDataLine line = (TargetDataLine) AudioSystem.getLine(info);
    line.open(format);
    line.start();
    AudioInputStream audio = new AudioInputStream(line);
    RecognizeWithWebsocketsOptions options = new RecognizeWithWebsocketsOptions.Builder().audio(audio).interimResults(true).timestamps(true).wordConfidence(true).contentType(HttpMediaType.AUDIO_RAW + ";rate=" + sampleRate).build();
    service.recognizeUsingWebSocket(options, new BaseRecognizeCallback() {

        @Override
        public void onTranscription(SpeechRecognitionResults speechResults) {
            System.out.println(speechResults);
        }
    });
    System.out.println("Listening to your voice for the next 30s...");
    Thread.sleep(30 * 1000);
    // closing the WebSockets underlying InputStream will close the WebSocket itself.
    line.stop();
    line.close();
    System.out.println("Fin.");
}

Example 5 with RecognizeWithWebsocketsOptions

use of com.ibm.watson.speech_to_text.v1.model.RecognizeWithWebsocketsOptions in project java-sdk by watson-developer-cloud.

the class RecognizeUsingWebSocketsWithSpeakerLabelsExample method main.

/**
 * The main method.
 *
 * @param args the arguments
 * @throws FileNotFoundException the file not found exception
 * @throws InterruptedException the interrupted exception
 */
public static void main(String[] args) throws FileNotFoundException, InterruptedException {
    FileInputStream audio = new FileInputStream("src/test/resources/speech_to_text/twospeakers.wav");
    Authenticator authenticator = new IamAuthenticator("<iam_api_key>");
    SpeechToText service = new SpeechToText(authenticator);
    RecognizeWithWebsocketsOptions options = new RecognizeWithWebsocketsOptions.Builder().audio(audio).interimResults(true).speakerLabels(true).model(RecognizeOptions.Model.EN_US_NARROWBANDMODEL).contentType(HttpMediaType.AUDIO_WAV).build();
    RecoTokens recoTokens = new RecoTokens();
    service.recognizeUsingWebSocket(options, new BaseRecognizeCallback() {

        @Override
        public void onTranscription(SpeechRecognitionResults speechResults) {
            recoTokens.add(speechResults);
        }

        @Override
        public void onDisconnected() {
            lock.countDown();
        }
    });
    lock.await(1, TimeUnit.MINUTES);
}