Mohamed Debyeche

Learn More
In this paper a new variant of HMM named distributed VQ HMM (DVQHMM) is presented. Its main characteristic is the use of a code books distributed on HMM states with a new manner of HMM parameters estimation. Procedures for training and HMM evaluation of each recognition unit are described. Comparative results on an isolated phoneme recognition system are(More)
This paper investigate the influence of GSMEFR speech Data on the performance of a text independent Speaker Identification System (SIS) based on Gaussian Mixture Models (GMM) classifiers. The performance evaluation due to the use of the GSMEFR speech Data, obtained by passing the local ARADIGIT database through the GSM coder/decoder. The recognition(More)
The paper presents an improved Vector Quantization (VQ) approach for discrete Hidden Markov Models (HMMs). This improved VQ approach performs an optimal distribution of VQ codebook components on HMM states. This technique, that we named the Distributed Vector Quantization (DVQ) of hidden Markov models, succeeds in unifying acoustic microstructure and(More)
This paper investigates the use of a Time Delay Neural Network (TDNN) as fuzzy vector quantizer to improve the Distributed scheme of HMM speech recognition. We investigate how to optimize the use of the Vector Quantization (VQ) by combining complementary preprocessing techniques based on multi-streams acoustic analysis. Then, in order to eliminate the(More)