Artificial intelligence (AI) is increasingly finding its way into all areas of everyday life. In media production and processing, AI is also one of the most important tools of the future for solving problems in an automated and thus cost-saving way.
Control tasks with variable and diverse input variables, such as the clever storage of media content in the cache of edge servers, can be optimized with self-learning algorithms. In addition, AI-based methods enable automatic content analysis, among other things for semi-automated production or for automated and immediate trend analysis of reactions to media content broadcast in social networks. An exemplary application, which is already being prototypically investigated in several public broadcasters, is AI-based facial recognition. The goal is to automatically recognize celebrities, such as politicians, but also presenters, in audio-visual contributions and to equip the associated contributions with corresponding metadata in order to find them more easily in the future.
The use of artificial intelligence in areas with a high "manual labour input", control algorithms, semi-automated production and preparation of content (e.g. robot journalism) and metadata, for an automated trend analysis of content in social media networks, an improved SDR/HDR conversion or object extractions in video sequences also appears to be particularly promising.
It is the task of IRT to examine the already existing algorithms with regard to their mode of operation, economic efficiency and reliability and to question their areas of application in terms of machine learning, such as content analyses and system optimizations. The knowledge gained will be applied to problems in the media sector in close cooperation with broadcasters for their concrete application. In addition, IRT is working on extending and supplementing existing approaches, among other things in cooperation.