Intelligibility Enhancement of Alveolar Fricative in Cleft Lip and Palate Speech

Abstract

Cleft lip and palate (CLP) is a congenital disorder which affects nasal and oral cavities. The most common speech production problems demonstrated by individuals with CLP are related to hypernasality, nasal air emission, and misarticu-lations. In this work, phonemic substitutions in CLP speech are analyzed and the problem of substitution of fricative/s/ to /sh/ is addressed. A signal processing based approach is proposed to correct misarticulation, which affects the intelligibility of CLP speech. Modification of misarticulation is carried out by transforming post-alveolar fricative /sh/ to alveolar fricative /s/ using spectral energy compression method. Here, misarticulation is characterized in terms of two attributes along with the prior information of stimuli. The first attribute is energy concentration of misarticulated fricative /s/ around 2-4kHz. The second attribute is energy ratio (energy in the 2 - 4 kHz band to 4–6 kHz band) value. From the evaluation results, it is observed that the perceptual quality of speech is improved after spectral energy compression. Further, the quality of the modified speech signal energy ratio is compared with the original speech signal. The perceptual and objective evaluation illustrates that the modified speech results in improved intelligibility compared to the original speech signal.