Universal Threshold Calculation for Fingerprinting Decoders using Mixture Models

Abstract

Collusion attacks on watermarked media copies are commonly countered by probabilistically generated fingerprinting codes and appropriate tracing algorithms. The latter calculates accusation scores representing the suspiciousness of the fingerprints. In a 'detect many' scenario a threshold decides which scores are associated to the colluders. This work proposes a universal method to calculate thresholds for different decoders solely with knowledge of the accusation scores from the actual attack. Applying mixture models on the scores, the threshold is set up satisfying the selected error probabilities. It is independent from the fingerprint generation and can be applied at any decoder. Also no knowledge about the number of attackers or their strategy is needed.

A Novel Attack Model for Collusion Secure Fingerprinting Codes

“…[8], and methods to calculate the accusation threshold, e.g. [33], in order to get more insights of this new class of attacks. Beside the empirical evaluations, the theoretic background to this attacks is still void.…”

“…As such unfortunate cases are sought to be avoided, different approaches have been proposed, e.g. [16], [33], [19]. In this work we restrict ourselves to the threshold calculation based on rare event analysis proposed in [12] and [7], which was already applied in e.g.…”

“…One advantage of the 'detect one' accusation scenario is the independence of different threshold calculations (e.g. [36], [16], [33], [22]). This permits drawing valid conclusions on the attacks' effectiveness that cannot be made void by newly proposed thresholding techniques.…”