Get result of async operation to Google speech-to-text

[Bruno Leitão]

Hi all,

I perform a async request to google speech-to-text, and now i do not know how to get the result of operation.

I already know that operation was successful

Thanks

BL

[Nicholas (Google Cloud Support)]

Using speech.asyncrecognize as an example, here would be the workflow:

• Issue a POST request to https://speech.googleapis.com/v1beta1/speech:asyncrecognize as specified here.  Barring any exception, this will begin the operation and return an Operation.
  
• If the operation.done == true, you can get the results from operation.response or the error from operation.error in the event the operation failed.
  
• If the operation.done == false, you must poll for updates using the operations.get API providing the operation.name until operation.done == true
  

I hope this information and general guide is helpful.  Give it go and let me know if these behavior and values do not match with the documentation provided.

[Bruno Leitão]

I'm using the Google .Net API client. When i call: AsyncRecognize, the Operation returns only the parameter Name (e.g. 123456), others are null.

With Name parameter, i'm calling operations.Get and this is the return:

{

  "name": "123456",

  "metadata": {

    "@type": "type.googleapis.com/google.cloud.speech.v1beta1.AsyncRecognizeMetadata",

    "progressPercent": 100,

    "startTime": "2016-08-11T15:51:38.592569Z",

    "lastUpdateTime": "2016-08-11T15:51:40.323793Z"

  },

  "done": true,

  "response": {

    "@type": "type.googleapis.com/google.cloud.speech.v1beta1.AsyncRecognizeResponse"

  }

}

Any call to AsyncRecognize the Response parameter is null.

I need to know how can i get the result of my async operation.

[Aris Alexis]

I have the same problem with the response, it doesn't include any result. Does this mean it didn't understand anything from the audio file? I am sending a 40seconds linear16 raw PCM 16 bit signed file with clear voices. 

If there is a problem why isn't there an error reported?

[Nicholas (Google Cloud Support)]

Thank you for bringing this to our attention.  I have not been able to reproduce the results reported here.  When I tested a .wav PCM 16-bit little-endian signed recording, the result was exactly as described above.  I first received an operation name from the async API call and eventually got a response JSON with the Operation GET API that included transcribed text.

This may be be an issue with the audio file, language recognition, or something else specific to your use.  Please submit a new issue on our Google Cloud Platform public issue tracker and we can investigate this further.  Be sure to attach the file used and the AudioConfig sent with the recording so that we may attempt to reproduce your results.  Also, please link to it from here and back so that others may follow.

Thanks in advance for your patience.

On Wednesday, August 10, 2016 at 8:30:55 PM UTC-4, Bruno Leitão wrote:

[Ohad Perry]

happened in both mp3 and flac files.

getting 

{

  "name": "851885438749752509",

  "metadata": {

    "@type": "type.googleapis.com/google.cloud.speech.v1beta1.AsyncRecognizeMetadata",

    "progressPercent": 100,

    "startTime": "2016-09-10T20:58:37.142077Z",

    "lastUpdateTime": "2016-09-10T20:59:19.996062Z"

  },

  "done": true,

  "response": {

    "@type": "type.googleapis.com/google.cloud.speech.v1beta1.AsyncRecognizeResponse"

  }

}

and nothing else..

[Ohad Perry]

+1 happens to me as well. the audio is in hebrew and it's an mp3 file

[Dave C]

FYI, I experienced this same issue with LINEAR16 PCM data and setting the sampleRate to 16k.  I then realized that my actual data was at 44.1k.  After resampling, I get a reasonable response, although an error message or even an empty list of responses would be nice in the case that the recognizer didn't properly understand the audio.

The response for my admittedly malformed request was this:

{"metadata": {"@type": "type.googleapis.com/google.cloud.speech.v1beta1.AsyncRecognizeMetadata", "progressPercent": 100, "lastUpdateTime": "2016-09-29T21:57:16.761811Z", "startTime": "2016-09-29T21:56:56.355536Z"}, "response": {"@type": "type.googleapis.com/google.cloud.speech.v1beta1.AsyncRecognizeResponse"}, "name": "1622634151588923227", "done": true}

I'm not sure if it's in the docs, but I take this to mean that the absence of results field in the response indicates that the speech api did not recognize the audio.

On Wednesday, August 10, 2016 at 8:30:55 PM UTC-4, Bruno Leitão wrote:

[Sam Nyavor]

I am also getting a response with no result and dont know how to access the result even if there was a result. How do i know if there are no results?