Paper Abstract and Keywords |
Presentation |
2016-03-02 15:40
[Poster Presentation]
Comparative Study on Robustness of Synchronization Information Embedded into an Audio Watermarked Frame Jessada Karnjana, Pham Nhien, Shengbei Wang, Nhut Minh Ngo, Masashi Unoki (JAIST) EMM2015-83 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
Most of reported audio watermarking techniques are frame-based. That is, a host audio signal is segmented into frames, and then one watermark bit is embedded into one frame. Most of the reported techniques assume to know where each frame begins and ends in their extraction process. In practice, these frame positions might not be available. In that case, the extracted watermark can be damaged severely and easily by the cropping attack. This problem is known as the frame synchronization problem. Recently, there are many attempts at solving the problem. In this work, we preliminarily study one approach suggesting that not only the watermark is embedded into a host signal, but synchronization information is embedded as well. Least significant bits (LSB)-based and M-sequence methods are implemented and evaluated comparatively in terms of the robustness of the synchronization information and the sound quality of audio signal into which the synchronization code is embedded. Test results show that both methods can deliver correct frame positions with high accuracy and are robust against cropping attacks. Although the robustness of the M-sequence method is a bit better than that of the LSB-based method, the synchronization code embedded by both methods is fragile to signal processing attacks, such as band-pass filtering and MP3 compression. The different in sound quality between the original and the code-embedded signal, which is measured in terms of ODG, is greater than 􀀀1 for the LSB-based method when the number of replaced bit is 3. On the other hand, the M-sequence method introduces perceptual noise to the host signal. These results might suggest a common problem in the synchronization-code-embedding approach. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
frame synchronization / pseudo-random sequence / LSB modification / robustness of synchronization information / audio watermarking / / / |
Reference Info. |
IEICE Tech. Rep., vol. 115, no. 479, EMM2015-83, pp. 41-44, March 2016. |
Paper # |
EMM2015-83 |
Date of Issue |
2016-02-24 (EMM) |
ISSN |
Print edition: ISSN 0913-5685 Online edition: ISSN 2432-6380 |
Copyright and reproduction |
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
Download PDF |
EMM2015-83 |
|