Speech Recognizer Project API Guide
v
|
The Speech Recognizer Project provides a speech recognition library called MinHMM. The MinHMM Library implements a fixed-point low-resource speaker-dependent speech recognizer targeted toward embedded processors. The Library provides easy-to-use APIs that can voice enable applications. This document describes the MinHMM Libarary APIs. For information on using MinHMM consult the User's Guide.
The MinHMM Library uses an audio activity detection library, VAD, which is included with the Speech Recognition Project. The VAD Library can be used independently on its own to locate audio activity above background noise levels. Documentation for the VAD Library is included in this document.
The MinHMM Library software provides a comprehensive set of APIs for speaker-dependent speech recognition for a wide variety of embedded applications. MinHMM APIs allow an application to perform the following operations.
Obtain information about a model
MinHMM provides the size, name, amount of training, and whether a model is enabled, along with a count of the total number of models.
Run recognition search
The application can utilize MinHMM APIs to run the search continuously, or the search can be suspended after a phrase is recognized. The latter method can be used to implement a "push-to-talk" mode of operation.
Optional write/read model data to/from a file
A version of the library with this feature enabled is not currently supplied.
Using the MinHMM Library software requires operations outside the recognizer, such as allocating memory for models and processing, and collecting audio data in real-time to provide to the MinHMM APIs. The MinHMM User's Guide specifies the requirements that must be met. It is important to consult the User's Guide.
The MinHMM software includes example program source code that illustrates usage of MinHMM and provides code that can assist in setting up MinHMM properly.
For more information on the application, and available evaluation kits
See the download page.