Title: HMM-GMM based Amazigh speech recognition system

Authors: Safâa El Ouahabi; Mohamed Atounti; Mohamed Bellouki

Addresses: Laboratory of Applied Mathematics and Information System, Polydisciplinary Faculty of Nador, 62702 Selouane, Nador, Morocco ' Laboratory of Applied Mathematics and Information System, Polydisciplinary Faculty of Nador, 62702 Selouane, Nador, Morocco ' Laboratory of Applied Mathematics and Information System, Polydisciplinary Faculty of Nador, 62702 Selouane, Nador, Morocco

Abstract: This study presents conception and realisation of an automatic independent speech recognition system using hidden Markov model (HMM). The system recognises 33 letters in Amazigh language. System is found well performed and can identify the Amazigh spoken letters at 88, 44% recognition rate, which is well acceptable rate of accuracy for speech recognition. The tests were taken based on the HMM and Gaussian mixture distributions. Hidden Markov toolkit (HTK) has been used in implementation and test phases. The word error rate (WER) came initially to 29.41 and reduced to about 11.52% thanks to extensive testing and change of the recognition's parameters.

Keywords: ASR; Amazigh automatic speech recognition; HMM; hidden Markov model; GMMs; Gaussian mixture models; HTK; hidden Markov model toolkit.

DOI: 10.1504/IJSISE.2020.113564

International Journal of Signal and Imaging Systems Engineering, 2020 Vol.12 No.1/2, pp.47 - 53

Accepted: 13 Nov 2020
Published online: 11 Mar 2021 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article