Int. J. of Signal and Imaging Systems Engineering   »   2016 Vol.9, No.1

 

 

Title: Evolutionary structure of hidden Markov models for audio-visual Arabic speech recognition

 

Authors: Amina Makhlouf; Lilia Lazli; Bachir Bensaker

 

Addresses:
LRI (Laboratory of Computer Research), Department of Computer Science, University of Badji Mokhtar, BP. 12, Annaba, Algeria
LRI (Laboratory of Computer Research), Department of Computer Science, University of Badji Mokhtar, BP. 12, Annaba, Algeria
Department of Electronics, University of Badji Mokhtar, BP. 12, Annaba, Algeria

 

Abstract: In this paper, we present an Audio-Visual Automatic Speech Recognition System that combines the acoustic and the visual data. The proposed algorithm here, for modelling the multimodal data, is a Hidden Markov Model (HMM) hybridised with the Genetic Algorithm (GA) to determine its optimal structure. This algorithm is combined with the Baum-Welch algorithm, which allows an effective re-estimation of the probabilities of the HMM. Our experiments show the improvement in the performance of the most promising audio-visual system, based on the combination of GA/HMM model compared to the traditional HMM.

 

Keywords: automatic speech recognition; computer vision; HMM; hidden Markov models; GAs; genetic algorithms; hybrid models; signal processing; audio-visual fusion; AV fusion; Arabic; multimodal data modelling.

 

DOI: 10.1504/IJSISE.2016.074651

 

Int. J. of Signal and Imaging Systems Engineering, 2016 Vol.9, No.1, pp.55 - 66

 

Available online: 08 Feb 2016

 

 

Editors Full text accessPurchase this articleComment on this article