Read more
This book particularly presents basic and novel theories with practical examples of the two main topics: of Digital Processing of Acoustic Signals, such as : Speech and Heart Sounds. Namely : - Digital Room Acoustics Equalization, because in many sound reproduction systems the degradation quality of the speech signal is mainly caused by the room effect, and- Automatic Heart Sounds Segmentation, as it is an essential phase before objective analysis and diagnosis.Homomorphic standard method and its modified version for equalizing room acoustics effects on the speech signal are presented. An alternative approach to the Balanced Model Truncation method (standard method) is proposed. This when compared to the standard method allows reducing substantially the equalization filter order.A novel automatic heart sounds segmentation method is proposed and applied to professional regular rhythms heart sounds of adult patients. Recurrent Neural Network based segmentation is proposed as an alternative method for irregular heart sounds rhymes.This Book is intended for graduate students, lecturers and researchers who want to start-up a research path in such a discipline.
About the author
Maamar AHFIR is currently Professor at the University of Medea (Algeria) and Visiting Academic to the Applied DSP and VLSI Systems Research Lab at the University of Westminster London, UK. His research interests include: room acoustics, speech and heart sounds processing.
