Massive-scale video anti-piracy monitoring

The author's recent participation in the Small Business Innovative Research (SBIR) program has resulted in the development of a patent pending technology that enables the construction of very large and fast artificial neural networks. Through the use of UNICON's CogniMax^TM pattern recognition technology we believe that systems can be constructed that exploit the power of "exhaustive learning" for the benefit of certain types of challenging pattern recognition problems. The Viacom lawsuit against YouTube^TM in early 2007 brought to light the magnitude of the video piracy problem and caused us to examine the associated technical challenges to determine whether our technology might enable an effective solution. This paper presents a theoretical study that describes how a massive-scale anti-piracy video pattern recognition system might be constructed using a large/fast Radial Basis Function (RBF) artificial Neural Network (NN) to enable a solution. Several daunting technical challenges exist. First, the amount of copyrighted video content that has been generated over time and now must be protected is enormous. Second, the activity level that is generally present on a large file-sharing site such as YouTube presents any pattern recognition system with a torrent of video content to inspect. Third, the concept of "fair-use" implies that an anti-piracy policy is not simply based on identifying a few copyrighted video frames. To determine system feasibility, this paper derives a set of example requirements for such a system, lays out a hypothetical anti-piracy data processing architecture, and evaluates the performance of the example system configuration.