A comparative study of glottal source estimation techniques

Abstract

Source-tract decomposition (or glottal flow estimation) is one of the basic problems of speech processing. For this, several techniques have been proposed in the literature. However, studies comparing different approaches are almost nonexistent. Besides, experiments have been systematically performed either on synthetic speech or on sustained vowels. In this study we compare three of the main representative state-of-the-art methods of glottal flow estimation: closed-phase inverse filtering, iterative and adaptive inverse filtering, and mixed-phase decomposition. These techniques are first submitted to an objective assessment test on synthetic speech signals. Their sensitivity to various factors affecting the estimation quality, as well as their robustness to noise are studied. In a second experiment, their ability to label voice quality (tensed, modal, soft) is studied on a large corpus of real connected speech. It is shown that changes of voice quality are reflected by significant modifications in glottal feature distributions. Techniques based on the mixed-phase decomposition and on a closed-phase inverse filtering process turn out to give the best results on both clean synthetic and real speech signals. On the other hand, iterative and adaptive inverse filtering is recommended in noisy environments for its high robustness.

Breathy, Resonant, Pressed – Automatic Detection of Phonation Mode from Audio Recordings of Singing

“…This process is called inverse filtering [Fritzell 1992, Walker and Murphy 2007, Drugman et al 2012, Gudnason et al 2012. The vocal tract (throat, mouth and in some cases nose) forms the tube, which is characterized by its resonances.…”

Excitation modeling for HMM-based speech synthesis: Breaking down the impact of periodic and aperiodic components

“…Nonetheless, this approach requires the reliable and accurate separation of these components from each other using glottal inverse filtering, which is a difficult inverse problem [10,11]. In the second case, the filter corresponds to the overall spectral envelope of speech and the excitation is the residual signal obtained by feeding the speech signal through the inverse of the estimated filter.…”

Speech polarity determination: A comparative evaluation

“…This is especially true for GSGW whose error rate reaches 41% in the noisiest conditions. It is indeed known [6] that the performance of glottal flow estimation techniques rapidly degrades in such environments. Although the proposed OMPD method remains the best approach up to 20dB of SNR, it is clearly outperformed in more severe environments.…”

“…An error on its determination results in a severe impact on the reliability and accuracy performance. There are also some methods of glottal flow estimation and for its parameterization in the time domain which assume a positive speech polarity [6]. This paper proposes a new approach for the automatic detection of speech polarity which is based on the phase shift between two oscillating signals derived from the speech waveform.…”