A Framework for Online Semantic Adaptation of Scalable Video

Scalable video coding and content-based adaptation have become key technologies to adapt content to diverse constrained usage environments (such as PDAs, mobile phones and networks, ...), while keeping its semantics. In this paper is proposed a framework where the adaptation of scalable video is guided by some semantics in the content. Specifically, an adaptive skimming scheme guided by content activity and its use for fast browsing of sequences are described. Efficiency is achieved computing activity using compressed domain analysis. Several metrics are evaluated as measures of activity