Syllable modeling in continuous Speech recognition for tamil Language

Abstract

Automatic speech recognition transcribes spoken words into text newlineSome of the challenging issues in speech recognition include variability in newlinephonetic context inconsistency of speaking style unevenness among different newlinespeakers and noise in the environment In this thesis continuous speech newlinerecognition of the Tamil language is considered Tamil is an agglutinative newlinelanguage in which the morphological process of adding affixes to the base of newlinea word is very common Tamil is a syllable timed language which has a newlineunique feature called the prosodic syllable Intonation and duration of an newlineutterance in Tamil is based systematically on the prosodic syllable newlineTwo methodologies have been proposed to address the context newlinevariability among phonetic units newline newline