Design and Development of a Model for Converting Speech to Phoneme using Waveform Audio File Format

Abstract

A phoneme is generally the unit of a speech or a sound.  The phonemes are those which differentiate one word from the other and phoneme has its own importance in the field of linguistics.  The raw audio is recorded and normally it is converted to the speech directly using the Google APIs.  There are only few models and approaches to convert a raw audio to phoneme.  The accuracy part in the phoneme recognition is still a challenge.  In order to have better accuracy, a prototype design has been designed and the design is carried out as a prototype to have phoneme recognition with better accuracy.