home · mobile · calendar · bactac · 2001-2002 · 

BACTAC - Mulligan

Seeing People -- Vision Based Acquisition for Telepresence
Research Assistant Professor, Department of Computer Science

High speed desktop computers, digital cameras and Internet2 connections are making collaboration via immersive telepresence a real possibility. The missing link is currently the suite techniques to extract, transmit and render the information from cameras at a remote sight such that the local user has the compelling sense of "being there". Making these applications a reality requires real-time environment sensing to acquire compelling 3D scene models. For online multi-media applications such as Tele-Immersion we are concerned with the "immersive feeling" of the rendered models, including the quality of depth points in terms of accuracy and absence of outliers, and the responsiveness of the system necessary to communication and interaction among users. Studying the sources and impact of errors in our stereo systems will in turn help us understand and improve the design of our methods and systems.

The next step beyond high quality, high speed depth extraction is a level of modeling and prediction which will improve our extraction, transmission and rendering by integrating information acquired over time. In this talk I will describe a system I worked on at the University of Pennsylvania, some experimental evaluation work on trinocular stereo and my research directions here at CU.

Department of Computer Science
University of Colorado Boulder
Boulder, CO 80309-0430 USA
May 5, 2012 (14:24)