Abstract: In this paper, a new parameterization method sound discrimination of multimedia documents based on entropy phase is presented to facilitate indexing audio documents and speed up their searches in digital libraries or the retrieval of audio documents in the network, to detect speakers in purely judicial purposes and translate films into a specific language. There are four procedures of an indexing method are developed to solve these problems which are based on (parameterization, training, modeling and classification). In first step new temporal characteristics and descriptors are extracted. However, the GMM and SVM classifiers are associated with the other procedures. The MATLAB environment is the basis of the simulation of the proposed algorithm whose system performance is evaluated from a database consisting of music containing several segments of speech.

Debabi Turkia, Bousselmi Souha and Cherif Adnen, “Innovative Automatic Discrimination Multimedia Documents for Indexing using Hybrid GMM-SVM Method” International Journal of Advanced Computer Science and Applications(IJACSA), 10(1), 2019. http://dx.doi.org/10.14569/IJACSA.2019.0100136

@article{Turkia2019,
title = {Innovative Automatic Discrimination Multimedia Documents for Indexing using Hybrid GMM-SVM Method},
journal = {International Journal of Advanced Computer Science and Applications},
doi = {10.14569/IJACSA.2019.0100136},
url = {http://dx.doi.org/10.14569/IJACSA.2019.0100136},
year = {2019},
publisher = {The Science and Information Organization},
volume = {10},
number = {1},
author = {Debabi Turkia and Bousselmi Souha and Cherif Adnen}
}