Schwa Deletion in Hindi Language Speech Synthesis
Damodar Magdum1, Tejaswini Patil2, Maloji Suman3
1Damodar Magdum, Research Scholar, Koneru Lakshmaiah Education Foundation, Vaddeswaram (A.P), India.
2Tejaswini Patil, Principal Technical Officer, C-DAC, GIST, Pune (Maharashtra), India.
3Dr. Maloji Suman, Professor, Koneru Lakshmaiah Education Foundation, Vaddeswaram (A.P), India.
Manuscript received on 15 August 2019 | Revised Manuscript received on 27 August 2019 | Manuscript Published on 06 September 2019 | PP: 211-214 | Volume-8 Issue- 6S, August 2019 | Retrieval Number: F10420886S19/19©BEIESP | DOI: 10.35940/ijeat.F1042.0886S19
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Schwa deletion is important factor for conversion of Grapheme to Phoneme. In Hindi language each consonant has weak vowel. This weak vowel is called as inherent schwa. These schwa is deleted some cases in pronunciation. Written form and speech forms are different in Indian language. Schwa plays important role in speech form. Deletion and retention of weak vowel decides how words are pronounced. Words morphology is main factors that affects pronunciation. In current paper, we describe schwa handling, deletion and retention rules. Based on different rule we developed schwa deletion algorithm. This algorithm has been tested over 6000 high frequency words. We received accuracy result up to 80%. Based on result an application has been developed to provide user interface for the text processing component of text to speech system.
Keywords: Schwa, Grapheme To Phoneme, Weak Vowel, Inherent Schwa, Pronunciation.
Scope of the Article: Natural Language Processing