APSIPA Transactions on Signal and Information Processing > Vol 8 > Issue 1

Semi-fragile speech watermarking based on singular-spectrum analysis with CNN-based parameter estimation for tampering detection

Kasorn Galajit, Japan Advanced Institute of Science and Technology, Japan AND Thammasat University, Thailand AND NECTEC, National Science and Technology Development Agency, Thailand, kasorn.galajit@nectec.or.th , Jessada Karnjana, NECTEC, National Science and Technology Development Agency, Thailand, Masashi Unoki, Japan Advanced Institute of Science and Technology, Japan, Pakinee Aimmanee, Thammasat University, Thailand
 
Suggested Citation
Kasorn Galajit, Jessada Karnjana, Masashi Unoki and Pakinee Aimmanee (2019), "Semi-fragile speech watermarking based on singular-spectrum analysis with CNN-based parameter estimation for tampering detection", APSIPA Transactions on Signal and Information Processing: Vol. 8: No. 1, e11. http://dx.doi.org/10.1017/ATSIP.2019.4

Publication Date: 16 Apr 2019
© 2019 Kasorn Galajit, Jessada Karnjana, Masashi Unoki and Pakinee Aimmanee
 
Subjects
 
Keywords
Semi-fragilityDifferential evolutionSingular-spectrum analysisTampering detectionConvolutional neural network
 

Share

Open Access

This is published under the terms of the Creative Commons Attribution licence.

Downloaded: 1270 times

In this article:
I. INTRODUCTION 
II. PROPOSED WATERMARKING SCHEME 
III. CNN-BASED PARAMETER ESTIMATION 
IV. EVALUATION AND RESULTS 
V. DISCUSSION 
VI. CONCLUSION 

Abstract

A semi-fragile watermarking scheme is proposed in this paper for detecting tampering in speech signals. The scheme can effectively identify whether or not original signals have been tampered with by embedding hidden information into them. It is based on singular-spectrum analysis, where watermark bits are embedded into speech signals by modifying a part of the singular spectrum of a host signal. Convolutional neural network (CNN)-based parameter estimation is deployed to quickly and properly select the part of the singular spectrum to be modified so that it meets inaudibility and robustness requirements. Evaluation results show that CNN-based parameter estimation reduces the computational time of the scheme and also makes the scheme blind, i.e. we require only a watermarked signal in order to extract a hidden watermark. In addition, a semi-fragility property, which allows us to detect tampering in speech signals, is achieved. Moreover, due to the time efficiency of the CNN-based parameter estimation, the proposed scheme can be practically used in real-time applications.

DOI:10.1017/ATSIP.2019.4