A comparison of Gaussian Mixture Modeling (GMM) and Hidden Markov Modeling (HMM) based approaches for Automatic Phoneme Recognition in Kannada

Prashanth Kannadaguli, Vidya Bhat

Research output: Chapter in Book/Report/Conference proceedingConference contribution

9 Citations (Scopus)

Abstract

We build and compare phoneme recognition systems based on Gaussian Mixture Modeling (GMM) which is a static modeling scheme and Hidden Markov Modeling (HMM) which is a Dynamic modeling scheme. Both models were built by using Stochastic pattern recognition and Acoustic phonetic schemes to recognise phonemes. Since our native language is Kannada, a rich South Indian Language, we have used 15 Kannada phonemes to train and test these models. Since Mel-Frequency Cepstral Coefficients (MFCC) are well known Acoustic features of speech, we have used the same in speech feature extraction. Finally performance analysis of both models in terms of Phoneme Error Rate (PER) justifies the fact that Dynamic modeling yields better results over Static modeling and can be used in developing Automatic Speech Recognition systems.

Original languageEnglish
Title of host publication2015 International Conference on Signal Processing and Communication, ICSC 2015
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages257-260
Number of pages4
ISBN (Electronic)9781479967612
DOIs
Publication statusPublished - 01-01-2015
Externally publishedYes
Event2015 International Conference on Signal Processing and Communication, ICSC 2015 - Noida, India
Duration: 16-03-201518-03-2015

Conference

Conference2015 International Conference on Signal Processing and Communication, ICSC 2015
CountryIndia
CityNoida
Period16-03-1518-03-15

All Science Journal Classification (ASJC) codes

  • Computer Networks and Communications
  • Signal Processing

Fingerprint Dive into the research topics of 'A comparison of Gaussian Mixture Modeling (GMM) and Hidden Markov Modeling (HMM) based approaches for Automatic Phoneme Recognition in Kannada'. Together they form a unique fingerprint.

Cite this