Class: SpeechToTextV1

Add a corpus to a custom model.

Add a single custom word.

Add multiple custom words.

Creates a new empty custom voice model

Response looks like:

{
  "customization_id": "abc996ea-86ca-482e-b7ec-0f31c34e5ee9"
}

Replaces recognizeLive & friends with a single 2-way stream over websockets

Create a session Set-cookie header is returned with a cookie that must be used for each request using this session. The session expires after 15 minutes of inactivity.

Delete a corpus.

Delete a custom model.

Deletes the specified session.

Delete a custom word

Removing a custom word does not affect the custom model until you train the model with the Train a custom model method.

List corpora

Lists information about all corpora that have been added to the specified custom language model. The information includes the total number of words and out-of-vocabulary (OOV) words, name, and status of each corpus.

Example Result:

 {
     "corpora": [{
         "out_of_vocabulary_words": 1,
         "total_words": 233,
         "name": "corpus-1",
         "status": "analyzed"
     }, {
         "out_of_vocabulary_words": 3,
         "total_words": 20,
         "name": "corpus-2",
         "status": "being_processed"
     }]
 }

Get corpus details

Example response:

  {
    "name": "corpus-1",
    "total_words": 100,
    "out_of_vocabulary_words": 5,
    "status": "analyzed"
  }

Get customization details

Example response:

 { owner: '8a6f5bb1-5b2d-4a20-85a9-eaa421d25c88',
   base_model_name: 'en-US_BroadbandModel',
   customization_id: 'e695ad30-97c1-11e6-be92-bb627d4684b9',
   created: '2016-10-21T19:09:33.443Z',
   name: 'js-sdk-test-temporary',
   description: 'Temporary customization to test the JS SDK. Should be automatically deleted within a few minutes.',
   progress: 0,
   language: 'en-US',
   status: 'pending' }

List all customizations

Example response:

{ customizations:
    [ { owner: '8a6f5bb1-5b2d-4a20-85a9-eaa421d25c88',
        base_model_name: 'en-US_BroadbandModel',
        customization_id: '6a7785a0-9665-11e6-a73a-0da9193a4475',
        created: '2016-10-20T01:35:00.346Z',
        name: 'IEEE-test',
        description: '',
        progress: 0,
        language: 'en-US',
        status: 'pending' },
      { owner: '8a6f5bb1-5b2d-4a20-85a9-eaa421d25c88',
        base_model_name: 'en-US_BroadbandModel',
        customization_id: '9e2f6bb0-9665-11e6-a73a-0da9193a4475',
        created: '2016-10-20T01:36:27.115Z',
        name: 'IEEE-test',
        description: '',
        progress: 0,
        language: 'en-US',
        status: 'ready' },
      { owner: '8a6f5bb1-5b2d-4a20-85a9-eaa421d25c88',
        base_model_name: 'en-US_BroadbandModel',
        customization_id: '6b194e70-9666-11e6-a73a-0da9193a4475',
        created: '2016-10-20T01:42:10.903Z',
        name: 'IEEE-test',
        description: '',
        progress: 100,
        language: 'en-US',
        status: 'available' } ] }

Get information about a model based on the given model_id

List of models available.

Get the state of the engine to check if recognize is available. This is the way to check if the session is ready to accept a new recognition task. The returned state has to be 'initialized' to be able to do recognize POST.

Get a custom word

Lists information about a custom word from a custom language model.

Example output:

 {
    "sounds_like": ["N. C. A. A.","N. C. double A."],
    "display_as": "NCAA",
    "source": ["corpus3","user"]
 }

List all custom words

Lists information about all custom words from a custom language model. You can list all words from the custom model's words resource, only custom words that were added or modified by the user, or only OOV words that were extracted from corpora.

Example response:

{
    "words": [
       {
          "word": "hhonors",
          "sounds_like": ["hilton honors","h honors"],
          "display_as": "HHonors",
          "source": ["corpus1"]
       },
       {
          "word": "ieee",
          "sounds_like": ["i triple e"],
          "display_as": "IEEE",
          "source": ["corpus1","corpus2"]
       },
       {
          "word": "tomato",
          "sounds_like": ["tomatoh","tomayto"],
          "display_as": "",
          "source": ["user"]
       },
       {
          "word": "$75.00",
          "sounds_like": ["75 dollars"],
          "display_as": "",
          "source": ["user"],
          "error":" Numbers are not allowed in sounds-like"
       }
    ]
 }

Result observer for upcoming or ongoing recognition task in the session. This request has to be started before POST on recognize finishes, otherwise it waits for the next recognition.

Speech recognition for given audio using default model.

Creates a HTTP/HTTPS request to /recognize and keep the connection open. Sets 'Transfer-Encoding': 'chunked' and prepare the connection to send chunk data.

Reset a custom model.

Train a custom model.

Waits while corpora analysis status is 'being_processes', fires callback once the status is 'analyzed'

Note: the code will throw an error in case there in no corpus in the customization

Waits while a customization status is 'pending' or 'training', fires callback once the status is 'ready' or 'available'.

Note: the customization will remain in 'pending' status until at least one corpus is added. Calling this on a customization with no corpa will result in an error.

See http://www.ibm.com/watson/developercloud/speech-to-text/api/v1/#list_models for status details.