Lecture Notes in Speech Production, Speech Coding, and Speech Recognition

Mark Hasegawa-Johnson, University of Illinois at Urbana-Champaign

These lecture notes were written for a series of three courses (one undergraduate, two graduate) which I lectured or co-taught at UCLA in the spring of 1998. Since then, many folks have asked for copies, so in 2000 I decided to put them on my web page. This page has been edited little since then. The notes are generally out of date, and contain some known big errors, but might still be useful to you.

PDF Chapter
(click here) Table of Contents, Table of Figures
(click here) 1. Review of Basic DSP
(click here) 2. Speech Production
(click here) 3. Short-Time Signal Processing
(click here) 4. Linear Predictive Coding
(click here) 5. Spectral and Cepstral Distance Measures
(click here) 6. Engineering Models of Audition
(click here) 7. Speech Coding
(click here) 8. Speech Recognition
(Known big errors: material about what can and can't be computed by a two-layer network, Figs. 8.10 and 8.11, is wrong. A two-layer network can compute any function).
(click here) References
(Actually this is just a list of related textbooks)