A Novel Text to Speech Technique for Tamil Language using Hidden Markov Models (HMM)
A. Femina Jalin1, J. Jaya Kumari2

1A. Femina Jalin, Student, Department of Electronics and Communication Engineering, Noorul Islam Centre for Higher Education, India
2J. Jaya Kumari, Professor, Department of Electronics and Communication Engineering, Mar Baselios College of Engineering and Technology, (Kerala), India.

Manuscript received on 02 July 2019 | Revised Manuscript received on 05 July 2019 | Manuscript published on 30 August 2019 | PP: 38-47 | Volume-8 Issue-10, August 2019 | Retrieval Number: I8589078919/2019©BEIESP | DOI: 10.35940/ijitee.I8589.0881019
Open Access | Ethics and  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: Application of digital signal processing in speech processing plays a major part in our everyday life. Text to speech system lets people to see and read out loud consecutively. Text-to-speech synthesizers use synthesis techniques that require good quality speech. Text to speech conversion (TTS) can apply to many applications such as automation, audio recording and audio-based assistance system. Text to speech conversion can be applied for various multinational language as well as for a number of local languages. An efficient text to speech conversion for Tamil language with extreme accuracy is proposed in this work. Multi feature, with a Hidden Markov Model (HMM) predictor is used to convert text to speech efficiently. By using the proposed method, the precision of the framework is enhanced by a factor of 6% when contrasted with the traditional system. 
Keywords: Digital Signal Processing (DSP), Hidden Markov Model (HMM), Mean square error (MSE), Tamil Unicode, Text to speech conversion, (TTS).
Scope of the Article: Digital Signal Processing