Homeworks will be graded on both their correctness and their clarity. Homeworks containing many superfluous plots and pages will not receive full credit. Being clear and succinct is often as important as being correct. Make use of the Matlab 'subplot' command when possible to plot multiple images on the same page.

If you have any questions, please email the TA at holub@caltech.edu.

Homework 0

• We would like you to become familiar with Matlab if you have never used it before.
• This link may be a useful tutorial: http://www.math.ufl.edu/help/matlab-tutorial/
• After you have familiarized yourself with Matlab please take a look at the following code which shows how you can load in and manipulate images from the web. (homework0.m) (homework0_2.m)

Homework 1

• Location of 101 categories: http://www.vision.caltech.edu/feifeili/101_ObjectCategories/.
• You will have to modify or create any code not provided in the lectures (e.g. roc.m for problems 1 and 2).
• Homework 1 (Due Thurs, April 7)

Homework 2

• You should select a project of interest to you from the list of projects by Tues, April 12.
• You will need to read this paper by Belhumeur et. al, Eigenfacs vs Fisherfaces: Recognition Using Class Specific Linear Projection.
• Homework 2 (Due Thurs, April 14 before class).

Homework 3

• David Lowe paper which may be useful to you.
• Homework 3. (Due Tues, April 26 before class).

Project Descriptions