In parametric text-to-speech synthesis using Hidden Markov Model (HMM), the fundamental frequency (F0) parameter modelling is important because it has a direct effect on the prosody of synthetic speech. F0 is typically modelled by a discrete distribution for unvoiced speech and a continuous distribution for voiced, by using a multi-space distribution (MSD). However, F0 modelling using MSD-HMM i...

In the absence of voicing, the discrimination of "voiced" and "voiceless" stop consonants in whispered speech relies on such acoustic cues as burst duration and amplitude, and formant transition characteristics. The articulatory processes that generate these features of whispered speech remain speculative. This preliminary investigation examines the articulatory kinematics differences between w...

From a representational viewpoint, the “voiced” series of obstruents in Shanghai dialect can be specified in terms of complementary, contextconditioned tonal and segmental features: either low tone or glottal pulsing. Yet, some studies have proposed that, when the “voiced” obstruents can only be signaled by low tone (stress-initially), they retain something of segmental voicedness. This somewha...

This paper describes how formant frequencies of voiced and unvoiced speech can be predicted from mel-frequency cepstral coefficients (MFCC) vectors using maximum a posteriori (MAP) estimation within a hidden Markov model (HMM) framework. Gaussian mixture models (GMMs) are used to model the local joint density of MFCCs and formant frequencies. More localised prediction is achieved by modelling s...

Wigner Ville Transform is introduced as a suitable tool for the analysis of the speech signal, as it allows making a correct representation of the signal energy. First, we give a mathematical formulation of the Wigner Ville Transform of a stationary model of voiced speech. The speech model is expressed as a linear combination of the harmonic frequencies of fundamental frequency of the voiced sp...

In this study, the compensatory strategies speakers adopt for conveying the voiced-voiceless distinction of consonants in whispered speech were investigated. Of 8 native Dutch speaking subjects each, 26 Dutch minimal word pairs containing voiced and voiceless obstruents, were recorded. Acoustic analysis with the software Praat revealed that both in whisper and in normal speech (1) preceding vow...

This paper presents a text-independent speaker recognition system based on the voiced segments of the speech signal. The proposed system uses feedforward MLP classification with only a limited amount of training and testing data and gives a comparatively high accuracy. The techniques employed are: the Rasta-PLP speech analysis for parameter estimation, a feedforward MLP for voiced/unvoiced segm...

We investigated 2-month-old infants' perception of a subset of highly confusable English fricatives. In Experiment 1, infants discriminated modified natural tokens of the voiceless fricative pair [fa]/[oa] but only when the syllables included their frication noises. They also discriminated the voiced pair [va]/[oa] both with and without fricative noises. These results parallel those found with ...