Using a custom acoustic model for speech recognition

Acoustic model customization is available only for previous-generation models. It is not available for next-generation models.

Once you create and train your custom acoustic model, you can use it in speech recognition requests by using the acoustic_customization_id query parameter. By default, no custom acoustic model is used with a request. You can create multiple custom acoustic models for the same or different domains. But you can specify only one custom acoustic model at a time for a speech recognition request. You must issue the request with credentials for the instance of the service that owns the custom model.

A custom model can be used only with the base model for which it is created. If your custom model is based on a model other than the default, you must also specify that base model with the model query parameter. For more information, see Using the default model.

You can also specify a custom language model to be used with the request, which can increase transcription accuracy. For more information, see Using custom language and custom acoustic models for speech recognition.

Examples of using a custom acoustic model

The following examples show the use of a custom acoustic model with each speech recognition interface:

For the WebSocket interface, use the /v1/recognize method. The specified custom model is used for all requests that are sent over the connection.

var access_token = {access_token};
var wsURI = '{ws_url}/v1/recognize'
  + '?access_token=' + access_token
  + '&model=en-US_NarrowbandModel'
  + '&acoustic_customization_id={customization_id}';
var websocket = new WebSocket(wsURI);

For the synchronous HTTP interface, use the POST /v1/recognize method. The specified custom model is used for that request.

IBM Cloud

curl -X POST -u "apikey:{apikey}" \
--header "Content-Type: audio/flac" \
--data-binary @audio-file1.flac \
"{url}/v1/recognize?acoustic_customization_id={customization_id}"

IBM Cloud Pak for Data

curl -X POST \
--header "Authorization: Bearer {token}" \
--header "Content-Type: audio/flac" \
--data-binary @audio-file1.flac \
"{url}/v1/recognize?acoustic_customization_id={customization_id}"

For the asynchronous HTTP interface, use the POST /v1/recognitions method. The specified custom model is used for that request.

IBM Cloud

curl -X POST -u "apikey:{apikey}" \
--header "Content-Type: audio/flac" \
--data-binary @audio-file.flac \
"{url}/v1/recognitions?acoustic_customization_id={customization_id}"

IBM Cloud Pak for Data

curl -X POST \
--header "Authorization: Bearer {token}" \
--header "Content-Type: audio/flac" \
--data-binary @audio-file.flac \
"{url}/v1/recognitions?acoustic_customization_id={customization_id}"

You can omit the language model from the request if the custom model is based on the default model, en-US_BroadbandModel. Otherwise, you must use the model parameter to specify the base model, as shown for the WebSocket example. A custom model can be used only with the base model for which it is created.

Troubleshooting the use of custom acoustic models

If you apply a custom acoustic model to speech recognition but find that the quality of speech recognition does not improve, check for the following possible problems:

Make sure that you are correctly passing the customization ID to the recognition request as shown in the previous examples.
Make sure that the status of the custom model is available, meaning that it is fully trained and ready to use. For more information, see Listing custom acoustic models.