An engine for content-aware on-line video adaptation

In this paper we show a practical implementation of an adaptation engine based on content inspection of video material. The localization of domain independent hints or features that allow to infer the non homogeneous distribution of semantically relevant information, allows to dramatically reduce the amount of adapted data while maintaining the meaningful information. The extraction of these features is performed on-line, via techniques that operate on the compressed domain, following an abstraction model that allows transparent adaptation of DCT based video and wavelets based scalable video.