Course Description
Digital Speech Processing is a one term course (winter term) intended
for undergraduate students on the Faculty of Electrical Engineering
and Computing at the University of Zagreb.
Syllabus
Principles of speech production. Classifications of phonemes in
Croatian language, speech waveforms and spectrograms. Acoustic theory
of speech production, wave propagation in the vocal tract. Concatenated
lossless tube model and discrete time models. Linear prediction
in speech analysis, autocorrelation and covariance methods. Efficient
algorithms for calculation of the linear predictor. Short time analysis
and synthesis of the speech signal. Principles of voice coding,
VOCODER. Cepstral speech analysis for voicing and pitch estimation.
Automatic speech and speaker recognition.
Staff
Instructors is doc.dr.sc. Davor
Petrinoviæ . Teaching assistant is dr.sc. Davorka
Petrinoviæ.
|