EE368B - Image and Video Compression

Topics

• Introduction
• Some fundamental results of information theory
• Scalar quantization and vector quantization
• Human visual perception
• Predictive coding
• Transform coding
• Resolution pyramids and subband coding
• Interframe coding
• Motion estimation
• Motion compensated coding
• Coding standard (JPEG, H.261, H.263, and MPEG)

Grading

The course grade will be based on homeworks (25%), a peer review (25%), and a term project (50%). Homeworks (3-4 problem sets which require computer and Matlab) should be done individually, but discussions among students are encouraged. The term project will be done individually or in groups and should take about 40-50 hours in total per person. A project approval is required. The approval deadline is Nov. 3. Web submission of the project report is due to Dec. 1.. Please be aware that for the web submission no extensions are granted. Class-room presentations of the projects are scheduled for Dec. 1-8.

Emanation of Course Information

The class home page http://www.stanford.edu/class/ee368b contains all the course information, including lecture slides, assignments, and latest announcements. If you have any questions, you might find the answers in the home page. Otherwise, please direct generic questions about the course to the TAs or Kelly. Contact Prof. Girod or the TAs if you have questions about homeworks and projects.

The class email list is ee368b@lists.stanford.edu. Please make sure that you subscribe to the list, otherwise you might miss important announcements. You can subscribe by sending email to majordomo@lists.stanford.edu with the one line message "subscribe ee368b". To unsubscribe, do the same but replace "subscribe" with "unsubscribe". If you need any help, send an email to majordomo@lists.stanford.edu with the one line message "help".

The class newsgroup is su.class.ee368b. This is intended to be a platform for students to freely exchange information. We will NOT be checking the newsgroup regularly, so please make sure that you send your questions to the course personnel if you expect an answer from us.

Handouts

The lecture notes and the problem sets and occasionally other material will be made available in Adobe Portable Document Format (pdf) form. These can be read on virtually any platform using the free Adobe Acrobat Reader software.

Computing

We recommend that you use the Image Systems Engineering (ISE) Lab to do all your work in this class, although you can choose to stay with the Leland account or your own machine, with the only requirement that a copy of all your project files should be on ISE at the time of submission. We will create an ISE account for you in the first few weeks of class. Please watch out for our announcements regarding this issue. See the page Computing Resources for a description of the ISE Lab.