PITCH-ASYNCHRONOUS GLOTTAL INVERSE FILTERING OF NORMAL AND PATHOLOGICAL VOICES BASED ON HOMOMORPHIC PREDICTION

摘要

Inverse filtering of speech signals for the separation of vocal tract and glottal source effects has a wide variety of potential applications, including the assessment of glottis-related aspects of voice function. Among all existing approaches to inverse filtering, this paper focuses on homomorphic prediction. While not favoured much by researchers in recent literature, such an approach offers two advantages over others: it does not require previous estimation of the fundamental frequency and it does not rely on any assumptions about the spectral enevelope of the glottal signal. The performance of homomorphic prediction is herein assessed and compared to that of an adaptive inverse filtering method making use of synthetic voices produced with a biomechanical voice production model. The reported results indicate that the performance of inverse filtering based on homomorphic prediction is within the range of that of adaptive inverse filtering and, at the same time, it has a better behaviour when the spectral envelope of the glottal signal does not suit an all-pole model of predefined order.