Integrating AI Systems: Take-Home Lessons From The Intelligent Room Project

Michael H. Coen
AI Lab
MIT

Friday, November 13, 1998
11 a.m.
Fuller Labs 311

This talk discusses what we learned in the creation of our lab's Intelligent Room -- a highly responsive environment that was designed to experiment with new forms of natural, multimodal human-computer interaction. It is a space in which computation is used to enhance ordinary, non-computational activity, by connecting it with the real world and the human-centered events occurring within it.

Our approach has drawn from a wide variety of AI disciplines, including computer vision, natural language processing, machine learning, and agent architectures. In creating the Intelligent Room, we literally had to combine dozens of individual subsystems -- not only into a coherent whole, but more importantly, into an architecture that allowed their combined strengths to overcome their individual weaknesses.

This talk will present the Intelligent Room, design decisions encountered while creating it, and how our experiences with the Intelligent Room have shaped the creation of its successor.

Host

Carolina Ruiz

Colloquium Coordinator

Carolina Ruiz

Maintained by webmaster@wpi.edu
Last modified: Sep 27, 2006, 16:05 EDT