Prototyping Expressive 3D Social Worlds

3D virtual worlds are increasingly popular arenas for social interaction. There are new opportunities and possibilities for the style of communication but important aspects present in face-to-face meetings are absent in the artificial environment, including the visual cues of emotion that are provided by facial expressions, and a realistic representation of one’s geographical location. Our research attempts to enhance social interaction within virtual worlds by proposing a framework that would enable fully expressive Internet communication with the use of 3D expressive models. The framework contains three separate systems that support different aspects of social interaction within virtual worlds. Firstly, the Virtual City Maker creates a believable environment by automatically creating realistic identifiable geo-referenced 3D environments from a variety of aerial and GIS image data. Secondly, an automated 3D head modelling system provides a mechanism for generating and displaying expressions. Thirdly, the Emotion Analyser provides a mechanism for triggering the display of appropriate expressions by automatically identifying emotional words contained in text messages, the person to whom the emotional words refer to and the intensity. Each system in the framework is discussed with reference to their contribution to enhancing social interaction within 3D social worlds.