Beta
69016

Speech Watermarking using a Hybrid Strategy of both Empirical Mode Decomposition and Singular Value Decomposition

Article

Last updated: 25 Dec 2024

Subjects

-

Tags

-

Abstract

This paper presents a proposed robust speech watermarking approach. This approach aims to increase the speech watermarking robustness against different attacks. The method is based on Empirical Mode Decomposition (EMD) and Singular Value Decomposition (SVD). The speech signal is decomposed by EMD into its Intrinsic Mode Functions (IMFs), the first IMF transform to a 2-D format. The watermark image embedded into the singular values (SVs) of the first IMF. After watermark embedding, the speech signal transformed back into a1-D format. The first IMF preserves the speech signal perceptual quality, which leads to preserving the watermarked signal imperceptibility. The singular values matrix is stable against any small perturbation happens to the original signal which provide more secure and robustness against attacks. The proposed approach can be implemented on the speech signal as a whole or as a blocks. Block-based SVD implementation allows embedding more than one watermark in the speech signal which increase the opportunities and efficiency of watermark extraction in the presence of attacks. Simulation results show that using of EMD with SVD enhance the watermark extraction especially in the presence of attacks. A block-based implementation of the proposed speech watermarking also realize a higher correlation coefficient in the presence of attacks.

DOI

10.21608/mjeer.2020.69016

Keywords

Speech watermarking, EMD, SVD and IMF

Authors

First Name

Safaa

Last Name

el-Gazar

MiddleName

-

Affiliation

Dept. of Electrical and Communication Eng., Faculty of Electronic Engineering, Menoufia University, Egypt

Email

-

City

-

Orcid

-

First Name

Sami

Last Name

El-Dolil

MiddleName

-

Affiliation

Dept. of Electrical and Communication Eng., Faculty of Electronic Engineering, Menoufia University, Egypt.

Email

-

City

-

Orcid

-

First Name

Alaa M.

Last Name

Abbas

MiddleName

-

Affiliation

Dept. of Electrical and Communication Eng., Faculty of Electronic Engineering, Menoufia University, Egypt.

Email

-

City

-

Orcid

-

First Name

Moawad I.

Last Name

Dessouky

MiddleName

-

Affiliation

Dept. of Electrical and Communication Eng., Faculty of Electronic Engineering, Menoufia University, Egypt.

Email

-

City

-

Orcid

-

First Name

El-Sayed M.

Last Name

El-Rabaie

MiddleName

-

Affiliation

Dept. of Electrical and Communication Eng., Faculty of Electronic Engineering, Menoufia University, Egypt.

Email

-

City

-

Orcid

-

First Name

Ibrahim M.

Last Name

El-Dokany

MiddleName

-

Affiliation

Dept. of Electrical and Communication Eng., Faculty of Electronic Engineering, Menoufia University, Egypt.

Email

-

City

-

Orcid

-

First Name

Fathi E.

Last Name

Abd El-Samie

MiddleName

-

Affiliation

Dept. of Electrical and Communication Eng., Faculty of Electronic Engineering, Menoufia University, Egypt.

Email

-

City

-

Orcid

-

Volume

29

Article Issue

1

Related Issue

9963

Issue Date

2020-01-01

Receive Date

2020-01-23

Publish Date

2020-01-01

Page Start

39

Page End

49

Print ISSN

1687-1189

Online ISSN

2682-3535

Link

https://mjeer.journals.ekb.eg/article_69016.html

Detail API

https://mjeer.journals.ekb.eg/service?article_code=69016

Order

5

Type

Original Article

Type Code

1,088

Publication Type

Journal

Publication Title

Menoufia Journal of Electronic Engineering Research

Publication Link

https://mjeer.journals.ekb.eg/

MainTitle

Speech Watermarking using a Hybrid Strategy of both Empirical Mode Decomposition and Singular Value Decomposition

Details

Type

Article

Created At

22 Jan 2023