Article 6212

Title of the article

             FILTERING SPEECH SIGNALS USING THE METHOD OF MULTIPLE DECOMPOSITION                                     AND ENERGY EVALUATION OF EMPIRIC MODES

Authors

Alimuradov Alan Kazanferovich, Postgraduate student, Penza State University, alansapfir@mail.ru
Tychkov Alexander Yuryevich, Candidate of engineering sciences, assistant, sub-department of information-measuring technology, Penza State University, tychkov-a@mail.ru
Churakov Petr Pavlovich, Doctor of engineering sciences, professor, sub-department of information-measuring
technology, Penza State University, iit@pnzgu.ru 

Index UDK

621.391; 519.21 

Abstract

The article shows that the efficiency of speech signals processing directly depends on the noise eliminating quality. The authors suggest to use a method of empirical mode decomposition for speech signals noise-balancing. The researchers have designed an algorithm of noise reduction based on estimation of energy values of the separate empirical modes and their removing. 

Key words

speech signal, voice-activated control, filtering, ensemble empirical mode decomposition, energy density surface. 

Download PDF

 

Дата создания: 27.08.2014 08:19
Дата обновления: 27.08.2014 09:23