Speech Spectral Modeling and Speech Synthesis Using Finite Impulse Response Digital Filters
Date of Award
Fall 1982
Document Type
Thesis
Degree Name
Master of Science (MS)
Department
Computational Modeling & Simulation Engineering
Program/Concentration
Modeling and Simulation
Committee Director
Stephen Zahorian
Committee Member
John W. Stoughton
Committee Member
Meghanad Wagh
Call Number for Print
Special Collections LD4331.E53G67
Abstract
Synthetic speech can be produced using a discrete-time model of human speech production. This model depends on the use of digital filters to model the spectral characteristics of the vocal tract. Speech spectra can be efficiently encoded using a set of principal components basis vectors; however, these vectors are not in a convenient mathematical form for directly determining a digital filter for the speech production model. If discrete cosine basis vectors are used to approximate the principal components basis vectors, then the cosine basis vector coefficients can be used to determine a Finite Impulse Response(FIR) digital filter for speech spectral modeling and speech synthesis. Three FIR filter formulations are presented and the corresponding analysis/synthesis systems are implemented on a digital computer. These formulations vary in terms of phase, spectral amplitude and frequency scaling, and computational complexity. The FIR analysis/synthesis systems are compared with the analysis/synthesis systems based on the principal components basis vectors and linear predictor filters(Infinite Impulse Response filters). A sentence preference test is used to compare the various systems. The results of the tests indicate that FIR digital filters can be used to produce high quality synthetic speech.
Rights
In Copyright. URI: http://rightsstatements.org/vocab/InC/1.0/ This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).
DOI
10.25777/eqt4-z846
Recommended Citation
Gordy, Paul E..
"Speech Spectral Modeling and Speech Synthesis Using Finite Impulse Response Digital Filters"
(1982). Master of Science (MS), Thesis, Computational Modeling & Simulation Engineering, Old Dominion University, DOI: 10.25777/eqt4-z846
https://digitalcommons.odu.edu/msve_etds/89