Automatic determination of the frequency behaviour of the larynx tone by the method of linear prediction
Abstract
This paper presents a method of automatically tracing the frequency of the larynx tone, based on a numerical implementation of the method of linear prediction on a digital computer. It also discusses the algorithm for the practical implementation of this method, derived by the author, and presents the results obtained from the analysis of a continuous speech signal with a duration of approximately 2 seconds.References
[1]1 J. W. BAYLESS, S. J. CAMPANELLA, A. J. GOLDBERG, A survey of speech digitization techniques, Proceedings of Carnahan conference on electronic crime countermeasures, 80-81 (1972).
[2] A. DZIURNIKOWSKI, Microphonemes as fundamental segments of a speech wave, Primary segmentation — automatic searching for microphonemes, Advence Papers of the IV 1JCAI — 75, Vol. 2, Thilisi 1975.
[3] A. DZIURNIKOWSKI, Primary sequentation of speech sound signals in the SUSY system (in Polish), Reports JJ UW, 52 (1976).
[2] A. DZIURNIKOWSKI, Microphonemes as fundamental segments of a speech wave, Primary segmentation — automatic searching for microphonemes, Advence Papers of the IV 1JCAI — 75, Vol. 2, Thilisi 1975.
[3] A. DZIURNIKOWSKI, Primary sequentation of speech sound signals in the SUSY system (in Polish), Reports JJ UW, 52 (1976).
