A model-based approach for architecture-independent video decoding

Digital videos play an important role and are generally worth being preserved not only for historical reasons. In order to be able to play them, decoders are needed, which, however, as any kind of software are architecture-specific and can potentially not be executed on future hardware. To overcome this issue, we propose a multilevel approach that combines serialization standardization and model-based specification of video data. Both are architecture-independent and serve as a basis for the implementation of decoders on specific (future) hardware with a standardized access interface. This interoperability aspect simplifies the integration of video decoding capabilities on different architectures and the cooperative work on video data content.