Digital fingerprinting is an emerging technology for identifying users who have legitimate access to plaintext content but may use the content for unintended purposes, such as duplication and redistribution. For multimedia, fingerprints can be put into the content using embedding techniques that are typically concerned with robustness against a variety of attacks mounted by an individual. Ensuring the appropriate use of multimedia content, however, is no longer a security issue with a single adversary. The global nature of the Internet has brought media closer to both authorized users and adversaries. It is now easy for a group of users with differently marked versions of the same content to work together and collectively mount attacks against the fingerprints. These attacks, known as collusion attacks, provide adversaries a cost-effective method for removing an identifying fingerprint.
In this talk, tracing traitors using collusion-resistant fingerprinting for multimedia that jointly considers the encoding, embedding, and detection of fingerprints will be presented. A general formulation of fingerprint coding and modulation provides a unified framework covering orthogonal fingerprints, coded fingerprints, and other correlated fingerprints. Under this framework, we have proposed a new class of structured codes, known as Anti-Collusion Codes (ACC), and designed algorithms that allows for gathering forensic evidence of the guilt and for identifying colluders.