Content-Based Multimedia Identification: A New Ap- proach for Digital Rights Management

The explosive growth of multimedia content on the Internet is revolutionizing the way of multimedia security and copyright management. Unfortunately, the two major technological solutions in the past decade, namely encryption and watermarking, are ill-equipped to deal with the change. This article introduces the third solution, i.e., Content-based Multimedia Identification (CBMI). The key idea is to automatically generate a new descriptor from the content, called mediaprint, as the identifier of that media item. Compared with the active protection (i.e., encryption) and authentication (i.e., watermarking) approaches, CBMI offers a passive but reproducible and reliable approach to manage rights of the digital content.