16:332:527
Digital Speech Processing
Lecture, 3 Hours/week
Fall Semester 2003

Instructor:

Lawrence R. Rabiner
lrr@caip.rutgers.edu

Course Description:

This course covers the basic principles of digital speech processing, including a review of digital signal processing, the mechanics of speech production and perception, and the acoustic theory of speech production. Also covered are the basic techniques for speech processing in the time and frequency domains, including, short-time energy, short-time Fourier analysis, homomorphic methods, and linear predictive coding methods. The basic concepts of speech coding, including modern speech coders such as multi-pulse coding and code-excited linear prediction methods are also covered in this course. Finally we end with a set of lectures on text-to-speech synthesis, and speech recognition, with one complete lecture devoted to the hidden Markov model of speech recognition. Knowledge and proficiency with Matlab is required for this course. A Matlab-based term project will be required for all students taking this course for credit.

Textbook:

L. R. Rabiner and R. W. Schafer, Digital Processing of Speech Signals, Prentice-Hall Inc., 1978.

Grading:

Homework Term Project Mid-term Exam Final Exam

30 % 20 % 20 % 30 %

Prerequisites:

332:346 (Introduction to Digital Signal Processing), working knowledge of Matlab programming language.

Time and Location:

Tuesday-Thursday, 4:30pm-5:50 pm, CoRE Building Room 601.