Authors: Tapashi Kashyap Das; P.H. Talukdar; Kishore Dutta
Addresses: Department of Instrumentation and USIC, Gauhati University, Guwahati 781 014, India ' Department of Instrumentation and USIC, Gauhati University, Guwahati 781 014, India ' Department of Physics, Handique Girls' College, Guwahati-781 001, India
Abstract: We use linear prediction cepstrum coefficients (LPCC)-based features, namely, the weighted LPCC and delta weighted LPCC, to recognise Assamese vowel phonemes employing a discrete hidden Markov model (HMM). We create a small database for the Assamese vowels, spoken in isolation by 20 speakers with equal numbers of male and female speaker. Each spoken phoneme is repeated ten times by each speaker. Thus, our database consists of 1,600 phonemes out of which 1,000 phonemes are used for the training stage and the remaining 600 phonemes are used for the recognition stage in our experiment. The overall recognition rate of our experiment is nearly about 81.5%.
Keywords: linear prediction cepstrum coefficients; LPCC; Assamese vowel phonemes; phoneme recognition; hidden Markov model; HMM; cepstral features; Assamese vowels.
International Journal of Information and Communication Technology, 2015 Vol.7 No.2/3, pp.218 - 234
Received: 08 Jun 2013
Accepted: 19 Oct 2013
Published online: 25 Feb 2015 *