Speech enhancement using non-linear extended Kalman filter to train multi-layer perceptron with backpropagation

Tarek Mellahi; Adil Bouhous; Seif Allah El Mesloul Nasri

doi:10.1109/iceccme57830.2023.10252497

Speech enhancement using non-linear extended Kalman filter to train multi-layer perceptron with backpropagation

Tarek Mellahi, Adil Bouhous, Seif Allah El Mesloul Nasri

Research output: Chapter in Book/Published conference output › Conference publication

1 Citation (Scopus)

Abstract

The development of algorithms that enhance speech quality is crucial because of their wide use in various applications where noise-free signals are needed. As a contribution in this field to tackle the challenging problem of improving the quality of noisy speech, we propose in this paper an approach that deploys an extended Kalman filter (EKF) to train a multi-layer perceptron (MLP). Since speech damaged by noise is usually available, this approach is particularly useful for solving this issue. Experiments employing the NOIZEUS database demonstrate that the method we talked about above yields superior subjective and objective findings when compared to other optimization roads.

Original language	English
Title of host publication	2023 3rd International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME)
Publisher	IEEE
Number of pages	5
ISBN (Electronic)	9798350322972
ISBN (Print)	9798350322989
DOIs	https://doi.org/10.1109/iceccme57830.2023.10252497
Publication status	Published - 22 Sept 2023
Event	2023 3rd International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME) - Tenerife, Canary Islands, Spain Duration: 19 Jul 2023 → 21 Jul 2023

Conference

Conference	2023 3rd International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME)
Period	19/07/23 → 21/07/23

Keywords

Backpropagation
Electric potential
Mechatronics
Roads
Gaussian noise
Optimization methods
Speech enhancement

Access to Document

10.1109/iceccme57830.2023.10252497

Cite this

@inproceedings{9d649966cddd4a1ba119c08205644e32,

title = "Speech enhancement using non-linear extended Kalman filter to train multi-layer perceptron with backpropagation",

abstract = "The development of algorithms that enhance speech quality is crucial because of their wide use in various applications where noise-free signals are needed. As a contribution in this field to tackle the challenging problem of improving the quality of noisy speech, we propose in this paper an approach that deploys an extended Kalman filter (EKF) to train a multi-layer perceptron (MLP). Since speech damaged by noise is usually available, this approach is particularly useful for solving this issue. Experiments employing the NOIZEUS database demonstrate that the method we talked about above yields superior subjective and objective findings when compared to other optimization roads.",

keywords = "Backpropagation, Electric potential, Mechatronics, Roads, Gaussian noise, Optimization methods, Speech enhancement",

author = "Tarek Mellahi and Adil Bouhous and Nasri, {Seif Allah El Mesloul}",

year = "2023",

month = sep,

day = "22",

doi = "10.1109/iceccme57830.2023.10252497",

language = "English",

isbn = "9798350322989",

booktitle = "2023 3rd International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME)",

publisher = "IEEE",

address = "United States",

note = "2023 3rd International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME) ; Conference date: 19-07-2023 Through 21-07-2023",

}

Mellahi, T, Bouhous, A & Nasri, SAEM 2023, Speech enhancement using non-linear extended Kalman filter to train multi-layer perceptron with backpropagation. in 2023 3rd International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME). IEEE, 2023 3rd International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME), 19/07/23. https://doi.org/10.1109/iceccme57830.2023.10252497

Speech enhancement using non-linear extended Kalman filter to train multi-layer perceptron with backpropagation. / Mellahi, Tarek; Bouhous, Adil; Nasri, Seif Allah El Mesloul.
2023 3rd International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME). IEEE, 2023.

Research output: Chapter in Book/Published conference output › Conference publication

TY - GEN

T1 - Speech enhancement using non-linear extended Kalman filter to train multi-layer perceptron with backpropagation

AU - Mellahi, Tarek

AU - Bouhous, Adil

AU - Nasri, Seif Allah El Mesloul

PY - 2023/9/22

Y1 - 2023/9/22

N2 - The development of algorithms that enhance speech quality is crucial because of their wide use in various applications where noise-free signals are needed. As a contribution in this field to tackle the challenging problem of improving the quality of noisy speech, we propose in this paper an approach that deploys an extended Kalman filter (EKF) to train a multi-layer perceptron (MLP). Since speech damaged by noise is usually available, this approach is particularly useful for solving this issue. Experiments employing the NOIZEUS database demonstrate that the method we talked about above yields superior subjective and objective findings when compared to other optimization roads.

AB - The development of algorithms that enhance speech quality is crucial because of their wide use in various applications where noise-free signals are needed. As a contribution in this field to tackle the challenging problem of improving the quality of noisy speech, we propose in this paper an approach that deploys an extended Kalman filter (EKF) to train a multi-layer perceptron (MLP). Since speech damaged by noise is usually available, this approach is particularly useful for solving this issue. Experiments employing the NOIZEUS database demonstrate that the method we talked about above yields superior subjective and objective findings when compared to other optimization roads.

KW - Backpropagation

KW - Electric potential

KW - Mechatronics

KW - Roads

KW - Gaussian noise

KW - Optimization methods

KW - Speech enhancement

UR - https://ieeexplore.ieee.org/document/10252497

U2 - 10.1109/iceccme57830.2023.10252497

DO - 10.1109/iceccme57830.2023.10252497

M3 - Conference publication

SN - 9798350322989

BT - 2023 3rd International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME)

PB - IEEE

T2 - 2023 3rd International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME)

Y2 - 19 July 2023 through 21 July 2023

ER -

Speech enhancement using non-linear extended Kalman filter to train multi-layer perceptron with backpropagation

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this