Research Article

Recognition of Isolated Telugu Words with Soft and Hard Sounds using Hidden Markov Models

by  Shaik Shafee, B. Anuradha
journal cover
International Journal of Computer Applications
Foundation of Computer Science (FCS), NY, USA
Volume 98 - Issue 15
Published: July 2014
Authors: Shaik Shafee, B. Anuradha
10.5120/17257-7603
PDF

Shaik Shafee, B. Anuradha . Recognition of Isolated Telugu Words with Soft and Hard Sounds using Hidden Markov Models. International Journal of Computer Applications. 98, 15 (July 2014), 9-12. DOI=10.5120/17257-7603

                        @article{ 10.5120/17257-7603,
                        author  = { Shaik Shafee,B. Anuradha },
                        title   = { Recognition of Isolated Telugu Words with Soft and Hard Sounds using Hidden Markov Models },
                        journal = { International Journal of Computer Applications },
                        year    = { 2014 },
                        volume  = { 98 },
                        number  = { 15 },
                        pages   = { 9-12 },
                        doi     = { 10.5120/17257-7603 },
                        publisher = { Foundation of Computer Science (FCS), NY, USA }
                        }
                        %0 Journal Article
                        %D 2014
                        %A Shaik Shafee
                        %A B. Anuradha
                        %T Recognition of Isolated Telugu Words with Soft and Hard Sounds using Hidden Markov Models%T 
                        %J International Journal of Computer Applications
                        %V 98
                        %N 15
                        %P 9-12
                        %R 10.5120/17257-7603
                        %I Foundation of Computer Science (FCS), NY, USA
Abstract

In this paper an attempt has been made to recognize isolated Telugu words consisting of hard and soft sounds using Hidden Markov Models (HMM). In Telugu language ,the alphabets ka (? ), cha (? ), ta (? ), tha (? ), pa(? ) are called hard( paruSha- ????) and ga(?), ja(?), da(?), dha(?), ba(?) are called soft (saraLa- ???) consonants [1]. While recognizing such sounds, the chance of recognizing hard sounds as soft and soft sounds as hard is more as they are pronounced with almost equal structure of vocal activity. For Example, T (? ), may be recognized as D (?), and P (? ) as B (?). In case of Telugu word Thammudu means younger brother may be recognized as Dammudu which has different meaning. Different words of this kind being applied to HMM and analyze the success/error ratio using HMMs [2,4].

References
  • A Dictionary of the Mixed Dialects and foreign words used in Telugu by Charles Philip Brown -1854
  • Hidden Markov Models for Speech Recognition B. H. Juang; L. R. Rabiner ,Technometrics, Vol. 33, No. 3. (Aug. , 1991), pp. 251-272.
  • Modelling asynchrony in automatic speech recognition using loosely coupled hidden Markov models H. J. Nock, S. J. YoungCambridge University Engineering Department, Trumpington Street, Cambridge CB2 1PZ, UK,Accepted 22 March 2002
  • The Application of Hidden Markov Models in Speech Recognition by Mark Gales and Steve Young-Foundations and Trends in Signal Processing Vol. 1, No. 3 (2007) 195–304-(2008).
  • Speech Recognition using HMM: Performance evolution in noisy environment by Mikael Nilsson and Marcus Ejnarsson, Blekinge Institute of Technology,March 2002.
  • Speech Matlab exercises: IT University of Copenhagen Multimedia Technology Speech and IT Systems.
Index Terms
Computer Science
Information Sciences
No index terms available.
Keywords

Telugu Speech Isolated word Recognition Hard and Soft Sounds in Telugu Vector Quantization HMM and MATLAB.

Powered by PhDFocusTM