Automatic evaluation of hypernasality and speech intelligibility for children with cleft palate

Abstract

The speech of cleft palate (CP) patients has typical characteristics. Hypernasality and low speech intelligibility are the primary characteristics for CP speech. In this work, an automatic evaluation of different levels of hypernasality and speech intelligibility algorithm for CP speech was proposed, in order to provide an objective tool for speech therapist. To identify different levels of hypernasality, the short-time energy and Mel frequency cepstral coefficients were calculated as acoustic features, then Gaussian mixture model was applied as classifier. For the automatic speech intelligibility evaluation, the classical automatic isolated word recognition system was applied. The automatic speech recognition accuracy could be viewed as an indicator for various levels of speech intelligibility. The experiment results indicated that the proposed computer-based system achieved a good performance on the automatic classification of CP hypernasality and speech intelligibility levels. The average classification accuracy was over 79% for four types of hypernasality detection, and the automatic speech recognition accuracy decreased along with the drop of speech intelligibility.