EE 639 Advanced Topics in Signal Processing and Communication

Fall 2004: Multimedia Information System

[Home] [Syllabus] [Lectures] [Readings] [Homework] [Project Recommendations] [Final Projects] [Announcements]

EE639 Final Project Guidelines


Dates: Oct 8, 2004 in class.

In the one-page proposal, please identify topic, problem,
objectives, workplan, and references. Please be as specific as
possible. What reference papers you plan to review and critique?
What improvement do you plan to add to existing systems? What
new prototype/simulation do you plan to do? Keep in mind you
have 7 weeks to work on the project. The goal is to do something
interesting and we can all learn from everyone's work.

Final Project Presentation

Dates: Dec 17, 2004.

Every team is required to make a presentation in class. 10 mins for
each one-person team and 20 mins for each multi-person team. The time
limit will be firmly enforced. Please rehearse your talk in advance so
that you can prepare well and observe the time limit. The presentation
will be evaluated by your fellow classmates and myself.

Please send the presentation slides to me before your schedule presentation.

Final Report

Due: Dec 17, 2004, before presentations

The report should include a minimal length of 4 pages for each
1-person team and 8 pages for each multi-person team. The format
is single-spaced and single-column. Please adhere to one of the
following templates provided for you: Microsoft Word or Latex.
For survey type projects, please make sure to include your own explanations
or observations, instead of using the same materials from the reference
papers. Please summarize or compare findings from multiple sources of references.

For experiment type projects, please clearly describe what
algorithms/systems have been designed, implemented, and experimented.
A demonstration is required to show that the system is in proper
working condition. You may install such demo system on a website or
include them with your report (e.g., put the test data and demo system
on a CD ROM). Reports should include analysis data (e.g. performance
graphs), example results (e.g., search results), and your
observations. Don't include source codes in the report, but please
have them ready in case we ask for them.

Project Grading Policy:

The most important objective of the project is for you to gain more
in-depth understanding of selected topics of your interest. You
identify an extended topic related to what you have learned and done
in the homeworks, and define a reasonable work plan and complete
an individual or team study. At the end, you will teach the whole
class what you have learned in the final project!

Here is the breakdown of the importance of each component of your
project. The percentages are based on the overall course grade, so
they should add up to 60%.

As an incentive to encourage more truely "multimedia" projects, you can
drop your lowest-score homework if you attempt a project that handles
more than one media, for example, combining both audio and video or
video and text. Please indicate that in your proposal.

Suggested topics

A very good place to look for interesting projects are some of the
recent conferences. They include IEEE International Conference on
Image Processing (ICIP), ACM Multimedia, IEEE International Conference
on Multimedia and Expo (ICME), and IEEE International Conference on
Acoustics, Speech, and Signal Processing (ICASSP). The following
topics are suggested as examples only. You are encouraged to discuss
with me and create new topics.

Experimental Type:

1. Repeated patterns in broadcast television

The goal is to identify re-occurring patterns in broadcast television
such as television commercials and important news footages from
broadcast telvision for archival and monitoring. The project involves
the construction of a system that is capable of monitoring multiple
channels, capturing the video, extracting salient visual and audio
features, and matching with past patterns. I have recently started
this project in my group and you are strongly encouraged to talk with
me before starting this project. I have written a short abstract that
summarizes the motivation of this project.

2. Any-Time, Any-Place Television

The idea of watching television any time was started by the TV-AnyTime
Consortium. Its goal is to standardize the infrastructure for Digital
Video Recording (DVR) devices like TiVo. As mobile devices become more
and more popular, there is a growing demand for watching television
any time at your cell phone or PDA. In this project, you will build a
system that can stream live broadcast television to two hand-held iPAQ
PDA, provide various VCR-like functions for recording TV programs at
the server, as well as implement algorithms to allow users to skim
long video program in a short time. Again, please speak with me if you
want to work on this project.

3. Improvement of image retrieval system

This is an extension to your homework. Possible areas of exploration
include incorporation of object segmentation, spatial relationships,
better features, relevance feedback, active learning, and semantic

Survey Type:

1. 3-D shape retrieval systems

Summarize recent advances and discuss possible applications for 3-D
shape retrieval systems.

2. Experiential Computing

Experiential computing systems refer to the systems that capture,
store, and retrieve all aspects of personal experiences. The
nature of the information captured is substantially different from
typical archival data and its proper analysis and retrieval have many
open research questions. Summarize the works in various reserach
groups and identify potentials and limitations.

3. Applications in Surveillance Systems

Identifying different events and persons in surveillance data is one
of the most important research topics in this post Sept-11 era. Survey
recent advances in object tracking, event detection, face or other
biometrics recognition and their uses in surveillance systems.

4. Application in Life and Health Sciences

Identify specific topics in life and health sciences such as genomics,
medical imaging or disease-spreading modeling, where information
retrieval has played an important role. Survey the representative
techniques and discuss the differences in their applications compared
with traditional IR systems.

Feel free to propose your own topic!

Sen-ching Samson Cheung
Last modified: Tue Aug 24 11:20:57 EDT 2004