Text: Digital Coding of Waveforms, Jayant and Noll, Prentice Hall

Prerequisites: EEE407 and EEE350

Objective: The goal of this course is to introduce graduate students to the principles and applications of speech coding.

Outline:

The course will cover selected topics from the text book and from research papers: Characterizations of voiced/unvoiced speech,  analysis/synthesis models, performance and complexity issues - DSP Background: digital filters, -Waveform Coders: scalar/vector quantization, the ADPCM ITU (CCITT) G.726, sub-band and transform Coders, the ITU G.722  - sinusoidal transform coder (STC), multiband excitation coder (Inmarsat- DVSI IMBE)  - Vocoder  Methods: the channel, formant, and homomorphic vocoders, linear predictive vocoders, the LPC-10, mixed -excitation LPC - the new 2.4 kbit/s MELP federal standard,  RELP - Analysis-by-Synthesis LPC: multi-pulse excited LPC, the 9.6 kbits/s  Skyphone standard,  regular pulse excitation LPC, the GSM full- and half-rate standards, US-1, PCS 1900, code excited linear prediction, the IS-54 VSELP, the  FS 1016 CELP,  the G.728 low-delay CELP, the G.729 ACELP, the CDMA IS-96 QCELP and IS-127 EVRC, theIS-136,  the TrueSpeech G.723. Audio Coding, The MPEG and the AC-3 Algorithms

Course Coordinator: Andreas Spanias

Return to Signal Processing Courses