Instructor:
|
|
TA: Nam Vo
Office: CCB308 |
|
TA: Shray Bansal
Office: CCB308 |
|
********************************************************************************************
ALERTS for REGISTRATION:
August 20: There are
two sections. CS4495A is for UNDERGRADS.
There was supposed to be a cap of 40 –
as of Aug 20 there were 54. Overloads
anyone???? Usually many drop so we’ll
see if spots open. Then there is
CS4495GR is for GRADS. That had been by permission only but as of Monday was
available to MSCS students and as of Wed will be open to all grads. But it is capped at 20 for now.
August 21 3pm: The
restrictions on CS4495GR have been removed.
Any grad student
(including PhD) from any department should be able to
enroll. If you cannot
send me an email.
********************************************************************************************
Who is this for? This is a senior level undergraduate course for those interested in computer vision. It is also open to graduate students who need a solid undergraduate grounding before taking the graduate version (CS7495). Graduate students should enroll in CS6495.
Prerequisites
· Data structures you’ll be writing code that builds representation of images, features, and geometric constructions.
· A good working knowledge of Matlab or Python or if desperate C/C++ programming. The course will use Matlab in lecture demonstrations. Problem sets will be done in Matlab or Python unless you convince me to hand in C++.
· This course has more math than many CS courses: Linear algebra, vector calculus, and linear algebra (that is not a typo).
· No prior knowledge of vision is assumed though any experience with Signal Processing is helpful
Textbooks
The main book
will be the Ponce and Forsyth. Now that
there is a second edition, it covers more of the material that makes up this
course. The Szeliski
book is a great reference but really more of a modern review of the state of
the art methods – more appropriate for a graduate class.. There is a (legal!) on-line PDF of the Szeliski book so we will strongly recommend buying the
Forsyth and Ponce and using Szeliski's digital
version.
FP: Forsyth &
Ponce, Computer Vision: A Modern Approach (2nd Edition),
Prentice Hall, 2011, ISBN-10: 013608592X, ISBN-13: 978-0136085928 (on Amazon)
SZ: Richard Szeliski, Computer Vision:
Algorithms and Applications (book Web
site )
Administrative
· Web site: This site. Will have posted calendar/syllabus with posted slides, problem sets with data, other administrative stuff.
· T-square: The usual stuff. There is a web page under resources that points to this class web page.
· Slides: PDF slides will be posted by linking to the calendar. I will try to have them up before lecture. No guarantees. These slides are really the basis for the class and homework.
· Piazza: You will all receive an invitation to Piazza for CS4495. If not, send us email. This was invaluable for discussions of problem sets.
· Matlab access: if you don’t know how to get MATLAB access, first ask a friend. Then come see the TAs or me.
· Python access: some of you may use Python and OpenCV. We’ll give you some support but mostly you’re venturing into the great unknown.
Grading
Finalizing now. The grade is mostly based on a variety of problem sets. Currently scheduled to be 8, though the first one (project zero – we’re so CS) is really just to make sure you can read in an image, do something to it, and get some results. Problem sets are due Sunday at 11:59pm unless otherwise stipulated. The goal will be to have problem sets posted by Thursday, 10 days before they are due. There will also be an exam – we will see if it will be a final or a “near the end of the term” exam. This is just to test some fundamental understanding without a text or colleagues.
· Project 0: 5%
· Projects 1 – 7: 80%
· Exam – 15%
Late
Policy
Don't be late, it’s not fair to the TAs. We will not accept late home works for full
credit it unless you have (several day) prior approval from the professor for
really extenuating circumstances.
Submit what you have on time.
What we did in the
past and will do again this year: You
may submit revised/new sections of the problem sets late but with a 50%
reduction. More details on this later
but it does allow you to catch up if you fall behind. But the 50% penalty is fixed.
Honesty
and Integrity Policy
Projects are to be
done individually but you may collaborate at the white board level. You may
help each other with algorithms and general computation, but your code must be
your own. REPEAT: THE
CODE MUST BE YOUR OWN. ANYTHING TAKEN
FROM THE WEB IS NOT YOUR OWN!!!!
Last modified: 9/19/2013