Abstract: Michael Christel is one of the principal investigators on Carnegie Mellon University's Informedia Digital Video Library Project. Currently, Informedia contains over 1 Terabyte of automatically analyzed commercial, broadcast video. This talk will describe the technologies used in Informedia for extracting semantic information from video and audio. In particular, how Informedia is applying integrated speech, language, and image understanding techniques to create a full content, searchable video library. A new project, Informedia Experience On Demand is extending these techniques to unstructured, unedited video from sources such as surveillance cameras and personal camcorders. Informedia and an early prototype of Experience On Demand will both be demonstrated.
Christel received his Ph.D. from the Georgia Institute of Technology in the area of human computer interaction, focusing on practical implications for the development of multimedia interfaces for educational applications. His research interests include digital video libraries, information visualization, and computer support for situated learning.