A Text to Visual Speech Instant Messaging System

This paper describes the implementation of text-to-visual-speech instant messaging system using the Remote Method Invocation (RMI) and graphics functionality of Java, together with synthetic speech via the Microsoft Speech API. Our system allows users to communicate over a low-bandwidth network connection using text that is converted into a realistic talking face. The avatar of each user consists of a set of images created from a single photograph , using transformations based on average visemes (visual phonemes). These images are animated in time with an audio interpretation of the entered text to simulate a video stream of the other person talking.