Describing the Voice User Interface

SpeechStudio is the tool for describing the voice commands of the VUI and the actions these commands will perform in the application. These descriptions are maintained in a project-specific database called a SpeechStudio Project, normally as a component of a corresponding Microsoft Visual Studio (i.e., VB or VC++) Project.

SpeechStudio can be used to develop a VUI for new or existing applications. SpeechStudio supports VUI development for applications that have voice as the only user interface as well as applications where speech recognition is used to augment a Graphical User Interface (GUI). In addition to a wide range of facilities for describing what can be spoken to an application, SpeechStudio is also used to specify text-to-speech output and pre-recorded audio output.

Screenshot  Click to enlarge

The initial view of your application presented by SpeechStudio is a workspace window populated with three folders. Two of the folders, one for the menus and one for the dialogs (forms), contain images extracted from the application's existing graphical user interface (if any). The third folder is for grammars not associated with any GUI elements. In a GUI application, menu items and controls on dialogs link mouse and keyboard actions to the application: e.g., clicking a button, typing into an edit box, and scrolling a window are each connected to functions in the application. SpeechStudio has facilities to associate grammars with these menu and form (dialog) resources, making it a simple matter to speech-enable menu items and controls. If your application has no graphical resources, then the menu and form (dialog) folders will still be present, they'll just be empty.

The third type of application resource supported by SpeechStudio is standalone grammars not attached to any GUI items. Grammars describe phrases that can be spoken. SpeechStudio grammars also describe what actions should be performed when a phrase is recognized. Authoring grammars and incorporating them into the application is what SpeechStudio is all about.

SpeechStudio presents a visual interface to some of the functionality of the SpeechStudio control. As you construct grammars that describe the words and phrases a user can speak to your application you also select actions from a palette of routines implemented by the SpeechStudio control.

Grammars are usually not static. They contain variable items that are filled in when the application runs; names from the latest address book, specials from a seasonal catalogue, today's menu, etc. The SpeechStudio palette includes actions that retrieve phrases from the application to be incorporated in the grammars.

SpeechStudio is ideal for iterative development. It is easy to create new grammars and to change existing ones. Your initial grammars can be refined as you gain experience with the voice interface.

Next: How Does SpeechStudio Work?