-----------------------------------------------
Over the last decade, computers have moved from being simple data storage devices and calculating machines to become an everyday assistant in the lives of many people. Advances in wireless communications coupled with more powerful, portable and cheaper embedded devices have resulted in an increasing demand for human-centered computer architectures that can pervasively sense and recognize human interactions. In such architectures, the interaction with computers will extend beyond the confines of the desktop model. Thus, in order that future access to computing be natural, the environment itself will need to maintain an awareness and perception of the users with whom it is interacting, and be capable of unobstructed, transparent and intelligent interaction to the user's needs and commands. Computers will have to interact and engage in dialogue with the real world surrounding them in ways similar to the way we people do. The interfaces for such ! computers will not be restricted to just menus or haptic interaction by keyboard and mouse, but include a combination of speech, gestures, context and emotions where technology adapts to the user and not vice versa. I my talk I will illustrate an approach to natural 3D multimodal interaction in immersive environments that accounts for the uncertain nature of the information sources. The approach builds in part on prior research on disambiguating the user's intent in 2D and 2.5D user interfaces by fusing symbolic and statistical information from a set of 3D gesture, spoken language, and referential agents. I will outline the MAVEN (Multimodal Augmented and Virtual Reality Environment for Natural Interaction) architecture, and provide examples from a 3D multi-modal testbed, which was developed to explore the ideas behind the system development.
The MAVEN architecture is a collaboration with David McGee at Natural Interaction Systems LLC, Phil Cohen and Ed Kaiser at the Center for Human-Computer Communication, Oregon Graduate Institute for Science and Technology, Alex Olwal, Hrvoje Benko and Steven Feiner at Computer Graphics and User Interfaces Lab, Columbia University.
-----------------------------------------
Andrea Corradini received his Ph.D. in computer science from the Department of Neuroinformatics at the Technical University of Ilmenau, Germany, where he was awarded a Marie Curie Research Fellowship by the European Union. Dr. Corradini has also studied mathematics at the University of Trento, Italy, and worked at the firm, MediSYS GmbH, in Ilmenau, Germany, on a project founded by the AT&Q Leonardo European Community program. Upon moving to the United States, Dr. Corradini held first the position of postdoctoral fellow at the Center for Human-Computer Communication at the Oregon Graduate Institute of Science and Technology, Oregon, and later that of senior research associate. As of January 2003, he has been working as Assistant Professor at the Natural Interactive Systems Laboratory at the University of Southern Denmark, Odense, Denmark. His research interests include multimodal interaction, gesture recognition and analysis, natural language processi! ng, dialogue, and sensor fusion.