Research shows that seeing another person when talking to them does little to improve conversation, but seeing the objects being talked about helps a great deal. In various ways, the shared visual environment leads to the development of common ground--the detailed mutual understanding among participants about what each knows at any moment. The primary goals of this research are to understand more completely the role of shared visual spaces in both face-to-face and mediated settings, to discover how the usefulness of visual information interacts with tasks and participants, and to identify ways in which video-based systems can truly support remote collaborative work. The outcomes will be (1) theoretical understanding of the attributes of shared visual spaces that support conversational grounding and thus task performance, (2) detailed knowledge about the parameters that make a difference in visual communication systems for working on complex, collaborative tasks; and (3) technology that allows the accurate estimation of the focus of attention in distributed settings. If successful, the research will provide a scientific basis for the design of communication systems for such distributed activities as telemedicine, distributed education, and staffing of help desks for repair.
