Monday, 28 May 2018

How to get speakertag from google speech to text longRunningRecognize method?

This is the code i'm using
const speech = require('@google-cloud/speech');

const client = new speech.SpeechClient({
  projectId: Meteor.settings.private.googleCloud.projectId,
  keyFilename: getFilePath('google-cloud.json')
});

const config = {
  "enableWordTimeOffsets": true,
  "encoding": "WAV",
  "languageCode": "en-US",
  "sampleRateHertz": 44100,
  "model": "video"
};

const audio = {
  uri: "gs://my-project-name/jeff_bezos_1_mono.wav",
};

const options = {
  config: config,
  audio: audio,
};

client
  .longRunningRecognize(options)
  .then(data => {
    const operation = data[0];
    console.log('got a promise representation', data);

    const errorHandler = err => {
      console.log(err);
      throw(err)
    }
    const completeHandler = longRRResponse => {
      console.log('**** response ****');
      console.log(JSON.stringify(longRRResponse, null, 2));
    }
    const progressHandler = (metadata, apiResponse) => {
      console.log('progress ', metadata);
    }
    operation.on('error', errorHandler)
    operation.on('complete', completeHandler)
    operation.on('progress', progressHandler)
  })
  .catch(err => {
    console.error('ERROR:', err);
    fs.unlink(name);
  });

this is the following response I got back
{
  "results": [
    {
      "alternatives": [
        {
          "words": [
            {
              "startTime": {
                "nanos": 100000000
              },
              "endTime": {
                "nanos": 700000000
              },
              "word": "your"
            },
            .
            .
            .
          ],
          "transcript": "your goal is to be the largest online and you are retailer in the world 
beyond that what's the goal for our mission is Earth's most customer-centric 
company and I know what that mean I'll give you an example",
          "confidence": 0.9520494341850281
        }
      ]
    },
}

I did not find any information about speakerTag in response or how to get back speakerTag information from API?
In google api-explorer for recognize there is option to select speakerTag in the fields section
enter image description here
Thanks!


from How to get speakertag from google speech to text longRunningRecognize method?

No comments:

Post a Comment