Dynamic voice user interface

Ease of use has always been one of the most important goals in Human Computer Interaction (HCI) research [1]. Since speech is the most natural method of communication for humans, researchers of HCI have recently focused on voice UIs. Even though considerable advances are made in Voice Recognition field of HCI, the ease of use is yet to be achieved.

The users have certain goals in mind while using a computer. To reach these ends, they use specific functionality of certain applications. With current interfaces, the users have to go through opening the application first to access the functionality they want. If they want to activate the functionality of another application, they either have to launch that application or change the active window. Thus there is no unified and easy way of working at a feature level.

Today, users are overwhelmed by not only the number of applications, but also the number of features a typical application has. Finding the needed functionality on the screen among other functionality is a time consuming task that hampers ease of use. Applications' features are presented on the screen even when they are not needed. It is not possible to customize the UI such that only the features that the user wants are available.

The design of our tool. Dynamic Voice User Interface (DV-UI) that is presented in this thesis, addresses these issues to create a unified, customizable voice user interface. DV-UI uses speech to text, speaker identification, and text to speech synthesis to provide an easy to use voice interface. It lets the user select any functionality from a set of applications and associate them with voice commands in real time. In this way, users can reach the functionality they need by a single voice command of user's choice without having to open an application or looking for the functionality on the screen. DV-UI provides a user-centric environment by allowing the user to add, modify, and delete voice commands and save these preferences under their voiceprints via speech. We expect that this approach of unification and customization will help the voice user interface to become mainstream way of HCI.