Second European Conference on Speech Communication and Technology

Genova, Italy
September 24-26, 1991


Hybrid Sinusoidal Modeling of Speech Without Voicing Decision

A. J. Abrantes (1), J. S. Marques (2), Isabel M. Trancoso (2)

(1) INESC/ISEL; (2) INESC/IST, Lisbon, Portugal

This paper presents a new approach to sinusoidal modeling of speech which avoids the use of a voicing detector. The proposed model represents the speech signal as a sum of sinusoids and bandpass random signals. The use of two different sets of basis functions increases the robustness of the model since there is no need to switch between techniques tailored to particular classes of sounds. The sinusoidal basis functions with harmonically related frequencies allow an accurate representation of the quasi-periodic structure of the speech signal. The bandpass random functions, on the other hand, are better suited for high quality representation of unvoiced speech sounds, since their bandwidth is larger than the bandwidth of an harmonic. The amplitudes of all the two sets of basis functions are simultaneously estimated by a least squares algorithm and the output speech signal is synthesized in the time domain by the superposition of all basis functions multiplied by their amplitudes. Preliminary tests confirm the better performance of the hybrid model for operation with noise-corrupted input speech, relative to classical sinusoidal models with a strong dependency on voiced/unvoiced decisions. Keywords: Speech Modeling, Sinusoidal Modeling, Coding

Full Paper

Bibliographic reference.  Abrantes, A. J. / Marques, J. S. / Trancoso, Isabel M. (1991): "Hybrid sinusoidal modeling of speech without voicing decision", In EUROSPEECH-1991, 231-234.