Download now Free registration required
Collusion attack is a cost-effective attack against digital fingerprint. To develop an efficient collusion-resistant fingerprint scheme, it is very important for the detector to study the behavior of the colluders and the performance of collusion attack. Although, several prior works have been proposed in the literature to analyze the performance of collusion attack, few effort has been made to explicitly study the relationship between risk, i.e., the probability of the colluders to be detected, and distortion of collusion attack. In this paper, the authors investigate the risk-distortion relationship of the linear video collusion attack with Gaussian fingerprint. They formulate the optimal linear collusion attack as an optimization problem, where the colluders try to minimize the distortion subject to a risk constraint.
- Format: PDF
- Size: 318.48 KB