HMM Based Kannada Speech Synthesis using Festvox
Sadashiva V Chakrasali1, K Indira2, Shashank B Sharma3, Srinivas N M4, Varun S S5
1Sadashiva V Chakrasali, Department Of Electronics And Communication, Ramaiah Institute Of Technology Bangalore, India.
2K Indira, Department Of Electronics And Communication, Ramaiah Institute Of Technology Bangalore, India.
3Shashank B Sharma, Department Of Electronics And Communication, Ramaiah Institute Of Technology Bangalore, India.
4Srinivas N M, Department Of Electronics And Communication, Ramaiah Institute Of Technology Bangalore, India.
5Varun S S, Department Of Electronics And Communication,Ramaiah Institute Of Technology Bangalore, India.
Manuscript received on 6 August 2019. | Revised Manuscript received on 13 August 2019. | Manuscript published on 30 September 2019. | PP: 2635-2639 | Volume-8 Issue-3 September 2019 | Retrieval Number: C4934098319/2019©BEIESP | DOI: 10.35940/ijrte.C4934.098319
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: The process which involves generation of human like voice by a machine is called speech synthe- sis. The developments in the fteld of speech synthesis is vast in international languages, but it is limited in Indian languages like Kannada. This work aims at de- velopment of such a system for Kannada language using Festival and Festvox. It is based on parametric analysis and models of speech features, particular to a language and speaker. The system is memoryless and dynamic, wherein only extracted features are stored but not recorded audio. The training process involves speech data acquisition, pre-processing, labelling using Baum- Welch Iteration, whereas testing process involves text analysis, text segmentation, speech synthesis and qual- ity enhancement using acoustic HMM model develop- ment. The quality of synthesis is 3.52 dB to 5.02 dB as measured by Mel-Cepstral Distortion (MCD) score.
Index Terms— Baum-Welch algorithm, Hidden Markov Model (HMM), Kannada, MCD score, Speech synthesis.
Scope of the Article: Vision-Based applications