Computer and Human Vision I, Fall 2009


Building image by Escher

Teaching Assistants:
Daniel Glasner and Michael Klots

Classes are held on Sundays, 14:00 - 16:00 at Ziskind 1.

To join the course mailing list click here.

Course Messages:
Feb 07   An example exam, intended to help prepare for the exam on Thursday.
Feb 04   Graded exercise 4 can be found in the "Computer Vision" mailbox.
Feb 01   The exam will take place on Thursday February 11th between 10:00 and 13:00 at Ziskind 1 (the same classroom the course was given in.) The exam is "closed material", that is, you are not allowed to bring your class notes or any other material such as books etc. to the exam. The fourth exercise will be returned before the exam (either on Thursday or on Sunday). We will place the returned exercises in the "computer vision" mailbox and send out an e-mail to let you know that they are there. Slides from the class on lighting have been posted on the website.
Jan 20   The course schedule has been updated as follows. On Jan 24th Prof. Ullman will give a class on object recognition. On Jan 31st Prof. Basri will give a class on lighting. This will be the final lecture in this semester's course (there will be no class on Feb 7.) A lecture on human vision will be given in the "Advanced Topics in Computer Vision" course in the spring semester. The schedule page has been updated to reflect these changes.
Jan 17   Exercise set number 4 has been posted on the course website. This is a theoretical exercise on stereo and structure from motion. The exercise should be prepared and submitted individually. Please submit by Monday Feb, 1st to the "computer vision" mailbox.
Jan 17   Slides from today's class (motion II) have been posted on the schedule page.
Jan 17   Handouts summarizing the two geometry classes are now available on the schedule page.
Jan 10   Slides from today's class (motion I) are now available on the schedule page.
Jan 3   Exercise set 3 has been posted on the course website. This is a programming exercise about stereo reconstruction using alpha-expansion. This exercise should be prepared and submitted in pairs or individually. Please submit by Sunday Jan, 17'th. The hard copy should be submitted to the "computer vision" mailbox and the files should be send by e-mail to Micahel Klots at Micahel-dot-Klots-atsign-weizmann-dot-ac-dot-il.
Dec 22   Notes from the Michael's tutorial on image deblurring that was given on Sunday and Monday are now available at the "schedule" page.
Dec 20   A couple of corrections in exercise 2:
1. Please re-download the files ex2_q3.mat and ex2_q4.mat (even if you've already done that earlier today).
2. In question 3 (c) it should read (multiply \eta by 0.2) and (multiply \eta by 10) instead of (multiply \sigma by 0.2) and (multiply \sigma by 10). The .pdf has been corrected accordingly.
3. We have posted a function on the exercise website called "getConvMat.m" this is a function that generates a convolution matrix from a kernel. Please use this function instead of using "convmtx2.m" as we originally suggested, it simplifies the exercise slightly and will give you better results on the image boundaries.
Dec 20   The follow-up for today's tutorial will be tomorrow, 1pm-2pm, Ziskind 261 (the seminar room on the 2nd floor)
Dec 15   Exercise set 2 has been posted. This is a programming exercise about deblurring using natural image priors. This exercise should be prepared and submitted in pairs or individually. Please submit by Thursday Dec, 31'st. The hard copy should be submitted to the "computer vision" mailbox and the files should be send by e-mail to Daniel Glasner at daniel-dot-glasner-atsign-weizmann-dot-ac-dot-il.
Dec 7   Next Sunday Dec. 13, the "2009 Israel Computer Vision Day" will be held at the Interdisciplinary Center (IDC), Hertzeliya. This is an annual meeting which brings together the Israeli computer vision community. During the day researchers will present some of their recent works. Those of you who are interested are invited. For schedule, abstracts, and driving directions please see: 2009 Israel Computer Vision Day
Dec 2   In question 5 of exercise 1 it suffices to explain the value of f at even coordinates.
Nov 30   Notes on deconvolution using natural image priors have been posted on the schedule page. See section 1 and section 2 (up to 2.1 including) for some of the material covered in the image reconstruction class.
Nov 25   Exercise set 1 has been posted. This is a theoretical exercise about image formation, Fourier transform and convolution. This exercise should be prepared and submitted individually. Please submit your answers by Thursday Dec, 10th to the ``computer vision" mailbox (mailboxes are located on the second floor.)
Nov 23   For those interested, a basic Matlab tutorial will be held this Wednesday, Nov 25, 4-6pm in the "aquarium", a.k.a room 13. It's the class with the pc's, behind the glass wall, that you pass on your way from where they serve pizza on Sundays to the sys.admin. people (Lea, Yoram, David and the rest).
Nov 23   Slides from the two last classes (Fourier I and II) are now available.
Nov 08   Slides from today's class are now available on the website (go to the schedule page and click on the PPT link).
If you want to be added to the course's mailing list please send an e-mail to Daniel Glasner at daniel-dot-glasner-atsign-weizmann-dot-ac-dot-il.
Oct 29   Welcome to the Fall 2009 course website, please check this page regularly for updates. The first class will take place on Nov 8'th at Ziskind 1.