This is a reading seminar focusing on recent development of statistical pattern recognition techniques that are promising for solving problems in video indexing and audio-visual content analysis. The goal is to get familiar with the state of the art knowledge, gain insights on solving practical video indexing/analysis problems, and acquire some hands-on experience through computer programming.

Topics will include:



The class will collectively review, critique, and experiment with each of the selected papers. In each week, 1-2 papers will be reviewed, discussed and demonstrated through programming examples. Each student will be assigned one paper for presentation and demonstration. For other papers, each student is asked to write brief comments about every paper on his/her class web site before and after discussion.

We will provide image/video data, feature extraction tools, and sample classification problems to be used for experimentation in this class.

In addition to reviewing/presenting a paper, each student needs to complete a course project with a topic that's related to the topics covered in the course. Team projects (no more than 2 people per team) are encouraged.

To maximize the chance of interaction of all participants, we will limit the class size. Auditing is allowed, but everyone will be assigned a paper and asked to do a term project.