Indian Journal of Science and Technology

Abstract

Speech synthesis is an artificial method of converting written text into machine generated speech. The Main Theme of this paper is to describe an overview of existing popular and advanced speech synthesizing techniques. Speech synthesis is a method of artificially generating the human speech from mathematical models and machines using some control parameters. The device which synthesizes is called as a synthesizer, the device which synthesizes is called as synthesizer and the synthesizer may be implemented as hardware or software. So for the research is not successful in generating a speech signal, usually researchers extract the speech parameters from the recorded speech and synthesize the original signal from it. A synthesizer can be viewed as Mathematical modeling of the vocal tract by extracting the acoustic/vocal features to produce the artificial generated speech output. The generated speech quality is described by its naturalness and intelligibility. The naturalness of synthesized speech gives an idea about how closely the generated output resembles the sounds produced by human speech production. Intelligibility describes how about the output can be understood by the listener after the perceiving it.

Keywords: HMM, Modelling, Speech Synthesis, TTS, Vocal Tract