COMP 4060/7950: Computer Vision, Fall 2012

Instructor: Yang Wang
Lectures: M/W 3:45pm-5:00pm, E2-304 EITC (note the new location)
Office hours: M 2:00pm-3:00pm, W 12:00pm-1:00pm, or by appointment
Google group:
Desire2Learn: (mainly for assignment submissions)


  • Sep 14: Starting next week, the class will be in E2-304.
  • Sep 7: (1) As discussed in class, the new lecture time will be 3:45pm-5:00pm on Mon/Wed starting next week. (2) Course website updated, first lecture slide uploaded.
  • Sep 7: The email setting of Google group is changed: now the post reply will only go to the author of the message. If you want your reply to reach everyone, you need to email
  • Sep 6: A Google group is created for this class. Invitations have been sent to emails on file in Aurora. Let me know if you prefer another email address and I will add it manually (or you can join the group yourself using your Gmail account). I will use it for general announcement. You can also use it to communite with other students in the class for course-related matters (e.g. when looking for a project partner).
  • The first class is on Sep 7 (Friday). After the first class, I like to move the lectures to Monday/Wednesday only, with no lectures on Friday. If you are registered in this course, please take a moment to fill out the following two surveys: survey 1 (this survey is now closed, thanks everyone for responding), survey 2. I will fix the lecture time and office hour after the first class.

    Course Overview

    Computer vision is the discipline of ``teaching machines to see''. The goal of computer vision is to make sense of photographs, videos, and other imagery. In this course, we will cover both basic and more cutting-edge materials in computer vision. The course will provide foundations for understanding the computer vision literature and for doing independent research in this field.


    Basic knowledge of linear algebra, calculus, probability. For undergraduate students, this means you should have taken all the MATH/STAT courses required for the CS major (MATH1300, MATH1500, STAT1000) with good grades. Ideally, you should also have taken a course in probability (e.g. STAT 2400). Previous exposures to image processing and/or machine learning are desirable, but not required. Some assignments involve programming in Matlab. Previous experience with Matlab is not required, although students are expected to learn the language.

    Students are also required to do a final project, so you need to have basic programming and algorithm skills to finish a project. For the final projects, students can use any programming languages they choose.

    Students from other departments are welcome to take the course, as long as you have the background listed above. Please talk to the instructor if you have questions about your readiness for the course. Depending on your status (department, grad or undergrad), you might need to do some paperwork in order to register. Please contact CS and your home department for details.


    No required textbooks. The following books are useful as references.

    Syllabus (aka ROASS document)

    Schedule and lecture slides

    Homework assignments: you will have 4 grace days over the semester for these assignments. Other than that, late assignments will not be accepted.

    Please fill out and sign the following honesty declaration form (COMP 4060, COMP 7950). Then submit it as a hard-copy to me (either in person, or slip under my office door). You only need to do it once for this course. Instructors are not allowed to begin grading until this form is received.

    • HW0: due 11:59pm on Sep 14 (This assignment is meant to be a warm-up exercise. It will not be graded and the submission is optional.)
    • HW1: due 11:59pm on Sep 19 (extened to Sep 21). You also need this file ( for this assignment.
    • HW2: due 11:59pm on Oct 5. You need this file ( for this assignment.
    • HW3: due 11:59pm on Oct 19. You need this file ( for this assignment.

    Reading summary/review

    Paper presentation (COMP7950 students only) (presentation schedule)


    • Proposal: due Nov 5 at 11:59pm (submit to the "proposal" dropbox folder in D2L)
    • Final report: due Dec 10 (extended to Dec 15) at 11:59pm (submit to the "project final report" dropbox folder in D2L)
    • Project presentation: Dec 12 at 1:00pm-5:00pm, E2-461 (project presentation schedule)
    • Weekly meeting sign-up sheet


  • Matlab information
  • Useful links


    I would like to thank many people at other institutions for making their teaching materials available. Much of the course design is borrowed from Derek Hoiem, Kristen Grauman, Svetlana Lazebnik, Rob Fergus, Fei-Fei Li, Greg Mori, Deva Ramanan, among others.