Multi-modal mathematics: Conveying math using synthetic speech and speech recognition

Abstract

Over the past decade, the notion of multi-modal access to technology has moved from the realms of science fiction to reality. It is not now unthinkable to communicate with a machine using voice recognition software, and to have the computers' response spoken in a voice comparable in quality to a human. This paper outlines methodologies for the verbal presentation of mathematical material to a user using prosodically enhanced synthetic speech. We will describe methodologies for generating a prosodic model of the "equation space". © Springer-Verlag 2004.