TY - GEN
T1 - Worst case attack on quantization based data hiding
AU - Liu, Ning
AU - Subbalakshmi, K. P.
PY - 2006
Y1 - 2006
N2 - Currently, most quantization based data hiding algorithms are built assuming specific distributions of attacks, such as additive white Gaussian noise (AWGN), uniform noise, and so on. In this paper, we prove that the worst case additive attack for quantization based data hiding is a 3δ function. We derive the expression for the probability of error (Pe) in terms of distortion compensation factor, α, and the attack distribution. By maximizing Pe with respect to the attack distribution, we get the optimal placement of the 3δ function. We then experimentally verify that the 3δ function is indeed the worst case attack for quantization based data hiding.
AB - Currently, most quantization based data hiding algorithms are built assuming specific distributions of attacks, such as additive white Gaussian noise (AWGN), uniform noise, and so on. In this paper, we prove that the worst case additive attack for quantization based data hiding is a 3δ function. We derive the expression for the probability of error (Pe) in terms of distortion compensation factor, α, and the attack distribution. By maximizing Pe with respect to the attack distribution, we get the optimal placement of the 3δ function. We then experimentally verify that the 3δ function is indeed the worst case attack for quantization based data hiding.
UR - http://www.scopus.com/inward/record.url?scp=46249118665&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=46249118665&partnerID=8YFLogxK
U2 - 10.1109/ISM.2006.161
DO - 10.1109/ISM.2006.161
M3 - Conference contribution
AN - SCOPUS:46249118665
SN - 0769527469
SN - 9780769527468
T3 - ISM 2006 - 8th IEEE International Symposium on Multimedia
SP - 679
EP - 684
BT - ISM 2006 - 8th IEEE International Symposium on Multimedia
T2 - ISM 2006 - 8th IEEE International Symposium on Multimedia
Y2 - 11 December 2006 through 13 December 2006
ER -