Use of digital media has exploded in the past few years. Increasingly easy access to digital media and increasingly powerful digital media manipulation tools have made multimedia security a very important issue. New applications of multimedia data such as streaming media and wireless multimedia also impose many new challenges for content security and digital rights management. Two complementary techniques have been developed to address content security and digital rights management. One is multimedia encryption/scrambling. The other is digital watermarking.
· Multimedia encryption/scrambling
Multimedia
Encryption/Scrambling
Access control is a central piece of an
intellectual property protection system.
One common method for access control is through encryption. We are
the first to recognize that, for many applications such as streaming media
application, it is required that content encryption (including key
distribution) provides error resiliency, scalability, network friendliness and
capability of performing signal processing directly on the encrypted bitstream, just as is required for unencrypted content
formats. In addition, for some applications with real-time constraint, such as
wireless multimedia streaming to low-power devices or digital cinema
applications, a particularly desirable feature is low processing overhead,
which can usually be achieved through selective encryption and/or some
lightweight encryption techniques. A “wholesale” encryption of the
entire content bitstream is usually not desirable or
even feasible.
To that end, we have developed a framework of
performing selective encryption and spatial/frequency shuffling of compressed
digital content that maintains syntax compliance to the compression format
after content has been secured. With such a joint design of encryption and
compression, we can achieve the various aforementioned desirable
functionalities, which is extremely helpful for multimedia delivery over error
prone channels with dynamic bandwidth such as wireless networks or the
Internet. This format compliant selective encryption/shuffling framework has been adopted into the MPEG-4 IPMP Final Proposed Draft
Amendment (FPDAM).
Journal Papers:
1. W. Zeng, J. Lan and X. Zhuang, “Security for Multimedia Adaptation: Architectures and Solutions”, IEEE MultiMedia Magazine, pp. 68-76, April-June 2006.
2. M. Ji, S. Shen, W. Zeng, et al, “MPEG-4 IPMP extension – for interoperable protection of multimedia content, “ EURASIP Journal on Applied Signal Processing, Special Issue on Multimedia Security and Rights Management, 2004:14, pp. 2201-2213, 2004. (click here for preprint in pdf format ).
3. J. Wen, M. Severa, W. Zeng, M. Luttrell and W. Jin, “A format compliant configurable encryption framework for access control of video”, IEEE Tran. Circuits & Systems for Video Technology, Special Issue on Wireless Video, pp. 545-557, June 2002. (click here for preprint in pdf format ) . ( 57 citations using Google Scholar search)
4.
Conference Papers:
1.
W. Zeng, J. Wen and M. Severa,
“Fast self-synchronous content scrambling by spatially shuffling codewords of compressed bitstreams,”
Proc. IEEE Inter. Conf. Image
Proc., Sept. 2002. (click here for
preprint in pdf format )
2.
3. J. Wen, M. Severa, W. Zeng, M. Luttrell and W. Jin, “A format compliant configurable encryption framework for access control of multimedia,” in Proc. IEEE Workshop on Multimedia Signal Processing, pp. 435-440, Oct. 2001. (click here for preprint in pdf format )
4. Wenjun
Zeng and Shawmin Lei, "Efficient frequency domain video scrambling for
content access control" (full paper, 20% acceptance rate, click here for paper in pdf format ) ACM Multimedia '99 ,
Digital watermarking techniques have been proposed to embed signatures in the multimedia data to identify the owner, the intended recipients, and to check the authenticity of the multimedia data. Applications include copyright protection, owner authentication, content authentication, tamper detection and data hiding. My interest in this area has been focused on the fundamental issues such as “perceptual capacity” of digital watermarks, protocols for specific applications, and visual optimization in digital watermarking.
One of the fundamental issues in digital image watermarking is the capacity of the watermarking channels. We address this fundamental issue from a perceptual point of view. We have developed some of the first perceptually based robust watermarking schemes that make use of human perceptual models to embed the strongest possible watermark signal while providing transparent quality. Our work is well known (see citations) in the field for its fundamental contribution to the understanding of the tradeoff between watermark robustness and perceptual transparency. This class of JND (Just Noticeable Differences) based image-adaptive watermarking algorithms intelligently adapts to the local characteristics of still image or video data. As a result, the embedded watermarks are as robust as possible to common signal processing and intentional attacks. See Journal Papers [4] and Conference Papers [9-10]. A related link of this work is http://www.bell-labs.com/org/1133/Research/ImageVideo/watermarking.html
In addition, we address the protocols that are necessary
for valid watermark detection for resolving rightful ownership of the
images. We identify some crucial requirements for such a valid detection. In
particular, we are the first to point out that, for the particular
application of resolving rightful ownership, it is crucial to require that the
original image not be directly involved in the watermark detection
process in order to prevent a false claim. A general framework for validly
detecting the invisible watermarks is then proposed. The optimal detection
strategy within the framework is derived. See Journal Papers [3] and Conference
Papers [7-8]. This work is also widely
referenced. See citation list 1 and
list 2.
It is noticed that in most advanced perceptual models, the JND is characterized as a non-linear function of local image features. The non-linearity in embedding introduces difficulties in the detector design. We are also the first to contribute to the optimum detection problem for the non-linearly embedded watermarks. Our work focuses on the self-masking JND model, which is the most widely-used in digital watermarking for its simplicity. We propose a novel approach to transform the test signal to a perceptually uniform domain and then performs Bayesian hypothesis testing in that domain. The derived detector is a generalized correlation detector (GCD, as opposed to the conventional, linear correlation detector, or LCD), which pre-process each observed sample before the linear correlation process. We further mathematically prove the optimality of our results, by showing the fact that a locally optimum detector must be in the form of GCD, and ours achieves the best error performance among all GCDs in the sense of Neyman-Pearson criterion. Simulation results demonstrate the superior performances of the proposed detector over the LCD. People who are interested are referred to the Journal Paper [1] and the Conference Papers [1][2] for details.
We also address various other visual optimization strategies in digital image watermarking. See Conference Papers [3-7]. In fact, these strategies are closely related to those that have been used for perceptual image compression.
Journal Papers:
1.
W. Liu, L. Dong and
2. X. Kang, W. Zeng, J. Huang, “A watermarking scheme based on multi-band wavelet and principal component analysis,” to appear in International Journal on Network Security, vol.6, no.2, pp.116–121, Mar. 2008.
3. W. Zeng and B. Liu, "A statistical watermark detection technique without using original images for resolving rightful ownerships of digital images," IEEE Trans. Image Processing, vol. 8, no. 11, pp. 1534-1548, Nov. 1999. (click here for abstract, and here for paper in pdf format.). (130 citations using Google Scholar search )
4. C. Podilchuk and W. Zeng, "Image-adaptive watermarking using visual models," invited paper, IEEE Journal on Selected Areas in Communications, vol. 16, No. 4, special issue on Copyright and Privacy Protection, pp. 525-539, May 1998. (click here for pdf file) (528 citations using Google Scholar search )
Conference
Papers:
1. W. Liu, L. Dong and W. Zeng, "Optimum Detection for Spread-Spectrum Watermarking that Employs Self-masking", IEEE International Conference on Image Processing, San Antonio, Sept. 2007.
2.
W. Liu, L. Dong, and
3.
Wenjun Zeng, "Visual optimization
strategies in digital image watermarking," invited talk, Digital Imaging Group
Technical Forum on Copyright Protection at PMA
(Photo Marketing Association) meeting,
4.
Wenjun Zeng, "Visual optimization in
digital image watermarking," invited talk, Workshop on Multimedia and
Security at ACM Multimedia '99 , Orlando,
Oct. 1999. (click here for paper in pdf format )
5.
Wenjun Zeng and Shawmin Lei, "Digital
watermarking in a perceptually normalized domain," (click here for paper in pdf format ) in Proc. 33rd Annual Asilomar
Conf. on Signals , Systems, and Computers, Oct. 1999.
6.
W. Zeng, B. Liu and S. Lei, "Extraction of multiresolution watermark images for resolving rightful
ownership," in IS&T/SPIE Symposium on Electronic Imageing: Science & Technology, vol. 3657, pp.
404-414, Jan. 1999. (click here for paper in pdf format
)
7.
W. Zeng, "Digital watermarking and data
hiding: technologies and applications," Invited Talk, Proc.
International Conference on Information Systems, Analysis and Synthesis,
1998 (Click here for postcript
file )
8.
9.
C. Podilchuk and W.
Zeng, (html version) "Perceptual
watermarking of still images," in Proc. The First IEEE Signal
Processing Society Workshop on Multimedia Signal Processing, June 1997,
10. C. Podilchuk and W. Zeng, "Digital image watermarking using visual models," Technical Memo, Multimedia Communication Lab, Lucent Technologies, Bell labs., Sept., 1996. Also in SPIE/IS&T Electronic Imaging'97: Human Vision and Electronic Imaging, Feb. 1997.
Data Hiding Applications
Data hiding provides an efficient way to convey side information that can be used for many different applications. For example, data hiding can be used to convey side information that helps to achieve better performance and additional functionalities in multimedia communication.
To our best knowledge, we are the first to suggest using data hiding for error concealment/resiliency purpose in a standard compliant way. In [1], we proposed to use data hiding to convey a R-D based significance score associated with an image block that is used to signify the effectiveness that the associated block can be interpolated based on its surrounding information. The significance score can later be used to assist real-time dynamic rate shaping for transmitting compressed video over time-varying channels.
Our on-going work also includes applying data hiding to other multimedia communication problems.
Journal Papers:
1.
W. Zeng, Resilient video transmission and
multimedia database applications, Ph.D Thesis,
Electrical Engineering Dept.,
Conference
Papers:
2. W. Zeng, “Spatial-temporal error concealment with side information for standard video codecs,” in IEEE International Conference on Multimedia and Expo, July 2003.
Last Updated :