The central design problem for meeting-capture technology is how to balance the needs of different individuals along the chain from capture to viewing. Meeting participants want to minimize effort and disruption during the meeting, while maximizing flexibility of venue and information sources captured. Meeting content owners want to streamline dissemination of material while maintaining managerial control. Consumers of meeting content want an experience comparable to authored multimedia, with excellent audiovisual fidelity, indexed playback, and full access to display data. This talk describes how these trade-offs were managed in the design of the Quindi Meeting Companion, a commercial PC-based meeting-capture system. A central design decision in the Meeting Companion was to take advantage of naturally occurring sources of digital data in the meeting environment as a primary source of indexing, rather than relying on automatic detection of audiovisual events. The talk will discuss the implications of that decision and also describe innovative uses of the system by collaborative workgroups.

Speaker Bio:

Stan Rosenschein has been active in research, teaching, and research-based ventures, with an emphasis on representation, reasoning, and perception in intelligent agents. Stan served as the Director of the Artificial Intelligence Center at SRI International and as Consulting Professor of Computer Science at Stanford University and has held research and teaching posts at NYU, the Technion, the Rand Corporation, and Hebrew University. He is currently CEO of Quindi, a collaboration tools company developing meeting-capture software. Prior to founding Quindi, Stan was the founder and CEO of Teleos Research, acquired by Autodesk in 1996. Stan holds a B.A. from Columbia University and a Ph.D. from University of Pennsylvania.