The amadaptadddata
module is deprecated in IDOL Speech Server version 11.5 and later. HPE recommends that you use the newer Deep Neural Network (DNN) techniques, rather than GMM-based acoustic models.
This module is still available for existing implementations, but it might be incompatible with new functionality. The module might be deleted in future.
The amadaptadddata
module accepts acoustic adaptation audio and transcript data, along with an existing acoustic model, and analyzes the data and how well the current model fits with the data. It creates an accumulator file that is subsequently used to update the acoustic model parameters.
Currently you can adapt only GMM models, not the newer, better-performing DNN acoustic models. Generally, GMM model adaptation produces models that are still inferior to the out-of-the-box performance of DNN models.
Mode | Input | Output | Description |
---|---|---|---|
The amadaptadddata module does not accept input streams or write to an output stream. All inputs and outputs are file based. |
Example:
output <- amadaptadddata(–, input)
Some of these parameters do not have an internal default value. You must either set the configuration parameter to a suitable value, or set the value in the corresponding action parameter.
|