Speech Recognition Controller Overview

This document provides an overview of the Speech Recognition Controller (SRC) component.

Purpose

The Speech Recognition Controller component provides speech recognition services for clients requiring speech recognition functionality. The SRC component comprises a client utility, a data handler and a custom commands library.

Required background

Automatic Speech Recognition (ASR) is a technology that enables users to use speech to interact with devices. ASR works by converting speech into a machine-readable form. Voice activated dialling is an example of an ASR application.

Architectural relationships

The ASR Client Utility is dependent on the Multimedia Framework. The MMF acts as the interface between the ASR Client Utility and the low level DevASR. A Speech Recognition Services (SRS) controller plugin is responsible for transporting speech recognition requests from the MMF to the DevASR. For playback, the DevASR implementation uses DevSound to play the recorded speech.

The relationship between the SRC component and other multimedia components is shown below.

Figure 1. SRC component architectural relationship

API summary

The SRC component includes the following APIs:

API Description

Automatic Speech Recognition Client Utility

Client side interface for ASR.

For more information, see Automatic Speech Recognition Client Utility Overview.

Automatic Speech Recognition Custom Commands

Custom commands for ASR.

For more information, see Automatic Speech Recognition Custom Commands Overview.

Automatic Speech Recognition Data Handling

Recognition data handling for ASR.

For more information, see Automatic Speech Recognition Data Handling Overview.

Typical uses

The SRC component is used for the following:

  • Recognising spoken input.

  • Training a new model.

  • Manipulating grammars, lexicons and models.