Modern work in computer vision uses a toolbox of sophisticated numerical and statistical algorithms. Some of these algorithms have been pioneered by computer vision researchers, while others were developed in other fields such as machine learning, theoretical computer science, statistics, physics, and applied mathematics. The purpose of this course is to cover the basics of several of these algorithms, including properties, data structures, and insights into using these techniques in practice. Nine invited speakers will give tutorials on these algorithms and we will explore their applications in computer vision. The speakers are all acknowledged experts, so it's a unique opportunity to learn this material from "the horse's mouth", from people who have a great deal of experience with, and in many cases helped develop, these algorithms. Each speaker will give two talks--a tutorial at UW, and a research talk at Microsoft. Programming assignments will be to implement a small number of these algorithms and to try them out on interesting problem domains.
The class will meet three times a week.
Talks at UW (CSE 577) will be Tuesdays and Fridays 1:30-2:50 in
AC 305, talks will be video-taped and posted online.
Talks at Microsoft (CSE 590CV) will be Monday
(with two exceptions) 10:30-noon
in building 113/1021. Van leaves promptly at 9:45am from the parking lot
at the north end of the HUB.
The speakers are: