Abstract

In this paper, we present the latest development of an emotion extraction engine used for real -time internet text communication. Real time expressive communication is important as it provides aspects of the visual clues that are present in face-to-face interaction not available in ordinary text-based communications. In former papers, we demonstrated a text-to-emotion engine that can analyse the emotional content present in a real-time chat environment and can deliver the emotional parameters necessary to invoke an appropriate expressive image. In this paper, we present a quick and user-friendly method to generate the necessary expressive images. Images can be generated for six emotion categories from one original neutral image. In each category, three different emotion intensities can be achieved. Users only need to provide a single default image, define six control points and two control shapes to generate all the images. This paper also presents the preliminary findings from a series of experiments that have been carried out to test the efficiency of our expressive image generator.