CS281B
Computer Vision and Image Analysis
Winter 2017

Textbooks:

None required. The following is a partial list of graduate-level books that cover "traditional" computer vision topics that might serve as good references.

• Computer Vision - Algorithms and Applications, Szeliski
• Multiple View Geometry in Computer Vision, 2nd Ed., Hartley and Zisserman
• Computer Vision: A Modern Approach, 2nd Ed., Forsyth and Ponce
• Three-Dimensional Computer Vision, Faugeras

• T. Hastie, R. Tibshirani, and J. H. Friedman, "The Elements of Statistical Learning", 2nd edition, Springer, 2009
• Richard O. Duda, Peter E. Hart, and David G.Stork, "Pattern Classification", second edition, Wiley-Interscience, 2001
• John Shawe-Taylor and Nello Cristianini, "Kernel Methods for Pattern Analysis", Cambridge University Press, 2004
• Tom M. Mitchell, "Machine Learning", McGraw-Hill, 1997

Topics:

This particular incarnation of CS281b will focus heavily on deep learning techniques for computer vision. Topics may be added and/or deleted depending on the available time and interest of students. Please consult CV Lecture Notes and ML, PR and ANN Lecture Notes pages for more details

Labs:

Final exam:

Prerequisites:

• There are no specific pre-requisites for this course. However, a certain degree of sophistication in math and programming, commensurate with the gradaute standing in computer science and engineering, is assumed. In particular, solid background in calculus, probability, and linear algebra is a must.
• Or CS/ECE 181B is not a prerequisite for this course.
• Proficiency in at least one high-level programming languages (e.g., Python) is assumed.

Grading Policy:

• There is no exam (midterm or final).
• Your grades in this class will be determined by (1) class attendance and discussion (20%), (2) programming assignments (40%) and (3) a class project (program + presentation) (40%).

General Class Policies:

• This class has a Web site at http://www.cs.ucsb.edu/~cs281b. You will fine useful information such as announcements, on-line copies of syllabus and class notes, and links to other interesting Web sites. It is your responsibility to check the Web site on a regular basis.
• It should be noted that the specification in the syllabus can still change. ECI staff has not yet been able to get the GPU-version of Tensorflow running in CSIL. There might be other alternatives (e.g., Amazon web servers), but the service may not be free and knowlege of systems/packages installation may be required. Most machine learning packages use a Python interface, which may or may not be familiar to some students. We will have a discussion of all these issues.