CS 4495/7495
Computer Vision

Fall 2001
College of Computing 102 
Tues & Thurs 12:00-1:30


Description

This course introduces techniques for Computer Vision. The course presumes a reasonable background in calculus and linear algebra; probability would be an asset as well. The course is really made up of three units. The first considers image formation and processing as is immediately relevant to computer vision (we will try not to cover material best left for an Image Processing course) The second unit covers fundamental computer vision, the science of recovering information about the world from images. This work includes shape recovery, object recognition, and video understanding. Finally we will also cover interesting applications and some of the particular research here at CoC - after all, we have to have some fun too.

Instructor

Aaron Bobick
Email: afb@cc.gatech.edu
Office: CoC Bldg (CCB) 218
Phone (bad idea): 404-894-8591
Office hours: 2-3pm; email is best

Teaching Assistant

Alexander Stoytchev
Email: saho@cc.gatech.edu
Phone: (404) 894-9311
Office Hours: Wed. 2:00-3:30 pm; or by appointment.
Location: CCB Picnic Area


Syllabus

The first draft of the syllabus for this course can be found here. As the course solidifies we will generate a lecture by lecture table.

The lectures for this class are being captured. To find them go to eClass index and click on Computer Vision. (soon to be there!)

Slides from Jim Rehg's lectures (in Zip Format).


Requirements, Collaboration, and Grading

The grades will be assessed as follows:
 

Problem Sets (not all PS are created equal)

50%

Mid-term

15%

Reading critiques ([almost] All or none!)

5%

Final project

20%

Class Participation

10%

There will be 4 or so bi-weekly problem sets that will involve some Matlab and hopefully some thinking. Collaboration on problem sets is encouraged at the "white board interaction" level. That is, share ideas and technical conversation, but write your own code. All problem sets should be in on time. One late problem set is accepted late (but before the next one is due) without excuse. After that, get prior permission.

There will be a short mid-term (one hour) just to make sure we're all on the same page.

There will be occasional readings assigend. You will also be given a single question to answer and will be expected to hand in a less than full page answer (a paragraph may often suffice). If you are missing more than two critiques you lose the whole 5%. I.e. we just really want you to read these papers!

Undergrads and grads will be graded on separate curves; more is expected from a graduate project than an undergraduate project.


Text

The text for this course is Image Processing, Analysis, and Machine Vision by Sonka, Hlavac, and Boyle. Our sources tell us it is in the bookstore. As you will see we have selected this book because of the topics it covers not how it covers the topics.

Some extra readings

There will be additional required readings. When they are available electronically you will be able to find them here:

Sept. 4: The Laplacian Pyramid as a Compact Imgage Code
 

Missing page 13 from Bergen and Hingorani's paper "Hierarchical Motion-based Frame Rate Conversion".


Problem sets

Problem sets will be posted here.




Midterm Sample Solutions


Help with Matlab, Unix and Web publishing