HMM and Concatenative Synthesis based Text-to-Speech Synthesis

A rapidly developing field of technological advances in computers is text-to-speech synthesis, which is crucial to many different types of interaction between humans and computers platforms. In this study, we’ve employed Hidden Markov Models (HMM) and concatenative synthesis were used in the unit selection in text-to-speech (TTS)synthesis. To assess the effectiveness of our system, we employed a “generic English” (Genglish) database, which is less complex and has a smaller lexicon than spoken English. Further, the system trains using classification and regression tree (CART) is used to calculate the likelihood of that a that a particular input text sequence would convert to a different phonetization given a set of contextual parameters. This work shows the importance and the significance of natural-sounding, comprehensible speech samples that are correct. We have utilized MATLAB TTSBOX, a toolkit for text to voice synthesis.