Welcome to the home page for the Media and Machines lab at Washington University in St. Louis. The lab was formed in the fall of 2000 around the arrival of several new faculty with overlapping interests. The core researchers have expertise in vision, graphics, and learning on robots. In addition to research in their own fields, the members are also exploring joint projects which fall into the overlap between two or more fields.

The structure of the Media and Machines lab is based on sharing, shared space, equipment, ideas, problems, and solutions. "Interdisciplinary" has become everyone's favorite buzz-word; a shared lab space encourages students and faculty to look beyond their own problems to see how their work relates to others', something that's easy to do when those other people are sitting in the seat next to you, not down the hallway.

The common thread among the researchers in the lab is the interface between the computer and the real world. This involves developing methods for capturing, representing, interacting with, and displaying data from the real world. An example of this is the non-lambertian capture project. This project begins with using vision techniques to capture models of real world objects for use in computer graphics. This provides a test bed for determining egomotion from non-lambertian surfaces. Eventually, the tools developed in the object capture context will be combined with robot-captured data to provide automatically generated walkthroughs.

In addition to the larger projects the researchers are also exploring problems in their individual areas. Some examples of these are:

Can you use periodic motion to calibrate your camera?
Can artist's composition skills be used to create more interesting graphics images?
Can a robot learn by example?
...

The Media and Machines lab also provides support for several projects on campus.

• Project Aria works with students to put objects in space and on Mars. Keith Bennett is the primary lead on this project and is supported by Bill Smart who is providing robot expertise.
• The robotics, vision, and beginning and advanced graphics courses also use the resources of the lab for access to equipment and expertise not available through the department's classrooms.