Text to Speech Synthesizer for Afaan Oromo Using Hidden Markov Model

This study explores the application of natural language processing techniques with text to speech synthesis for the Afaan Oromo language using the “Hidden Markov model” on 600 news datasets that were prepared in collaboration with linguists and experts of Afaan Oromo language.  Speech synthesizers are the most essential in helping impaired people, in the teaching and learning process, for telecommunications and industries. The dataset was tested on a hidden Markov model algorithm. The synthesiser has two core components: training and testing phases. In this study, the subjective Mean Opinion Score (MOS) and objective Mel Cepstral Distortion (MCD) evaluation techniques are used. The subjective results obtained using the mean opinion score (MOS) are 4.3 and 4.1 in terms of intelligibility and naturalness of the synthesised speech, respectively. The objective result obtained using the mean opinion score is 6.8 out of 8 that is encouraging.