MAT 201A Media Signal Processing (Fall, 2009)
Contacts
Instructor: Stephen Pope, office South Hall 4340F
Office Hours: Tuesdays 1:00 - 3:00 PM or by appt.
TA: Javier Villegas
Office Hours: TBD.
Time/Place
Lectures: Tues/Thurs.
3:00 - 5:00 PM
Discussion Group/Lab: Monday 10:00 - 11:50 AM
Lectures are held in Music 2215
(CREATE class room)
Downloads and Web Resources
Course Overview
The MAT 201A course is dedicated to introducing students to the multimedia digital signal processing methods. We will explore a range of topics from theoretical principals to practical considerations in multimedia (speech, audio, still images, and video) signal representation, synthesis, analysis and processing. The course assignments will consist of reading and homework tasks where students explore the concepts introduced in the lectures through concrete software applications using the MATLAB programming language and custom- developed SSUM framework.
The premise of MAT 201A is that the multimedia (MM) expert of the future will need to have a firm grasp of the mathematical foundations of MM data representation and to understand (and even be able to extend) the algorithms used in common MM signal synthesis/processing/analysis applications.
The recommended text book is Signal Processing First by McClelland, Schafer & Yoder; the course reader consists of selected readings from the signal processing and multimedia literature; the course web site includes many links as well as down-loadable example software in MATLAB. Students are expected to purchase a personal license to MATLAB.
Prerequisites
Students are expected to have graduate standing (or the instructor’s approval), experience with trigonometry, complex numbers, and algebra, and to be functionally proficient in some programming or scripting language. The majority of the course grade will be based upon a final project report, presentation, and demonstration, in which the students explore one of the lecture topics in the form of a multimedia software application in MATLAB.
Format
- 2 * 2 hours/week lectures
- 1 * 2 hours/week lab
- Homework programs, quizzes, development project, final
presentation
Topical Outline
1. Multimedia applications tour
- Image, video, sound
- Capture/logging, archiving, synthesis, transcoding
- How are apps written?
2. Math & MATLAB introduction
3. Data representations
- Image, sound data structures
- Signal sampling and quantization
- Signal magnitude and phase
- Other properties
4. Operations on signals
- Signal creation and manipulation
- Signal arithmetic and statistics
- Signal-processing functions and flow-charts
- Memory, difference equations and digital filters
- Autoregression and moving-average processes
5. Information theory of signals
- Applications
- Error correction
- Compression
- Data embedding and hiding
6. Transforms and mappings
- Time- and frequency-domain signals
- Spectral representations, Fourier transform
- Linear prediction, signal modeling by polynomials
7. Techniques in applications
- Image edge/object/patch detection
- Motion estimation in video
- Feature extraction from speech and music
- Gesture capture and analysis
8. Projects
- SSUM examples throughout
- Image processing, feature extraction
- Video processing, computer vision
- Audio feature extraction, onset detection
- Gestural input and gesture detection
- Capture, archival, distribution, transcoding
Links to MATLAB and DSP Tutorials
DSP Guru Links - http://dspguru.com/info/tutor/other2.htm
How do I learn DSP? - http://www.redcedar.com/learndsp.htm
101 Tutorial - http://101science.com/dsp.htm
DSP Guide - http://www.dspguide.com/pdfbook.htm
MATLAB Tutorial - http://users.ece.gatech.edu/bonnie/book/TUTORIAL/tutorial.html
Student Projects
