One year of AI4Media – The main achievements

The first year of the AI4media project has been successfully completed in August 2021, delivering important outcomes as we work to achieve our main objectives: delivering the next generation of core AI advances for the Media sector, enhancing AI training, and reimagining AI as a crucial beneficial enabling technology in the service of Society, Democracy and Media.

Our main achievements during this year can be summarised as follows:

• Novel research tools & methodologies have been developed in four core AI areas: a) new learning paradigms and distributed AI, b) trustworthy (explainable, robust, privacy-preserving, fair) AI, c) multimedia content creation and analysis, and d) AI in the service of citizens and society (disinformation detection, fair recommendation systems, political debate analysis, effects of content sharing in real-life, etc.). Our research has already resulted in more than 70 publications in prestigious venues, open-access software in GitHub, and 4 new open datasets;

• The user requirements for the seven AI4Media use cases have been identified through an intense co-creation process that involved strong collaboration between AI researchers & media industry partners. This process resulted in an impressive mapping of AI4Media use cases and the media industry’s AI requirements across the whole media value chain.

• A detailed overview and analysis of AI policy and regulatory initiatives on the EU level have been delivered as a first step for the establishment of the Media AI Observatory. The analysis covers a wide variety of policy initiatives, including political, ethics & trust, IPR, safety & liability, and AI-specific, while it also examines regulatory initiatives like the AI package, DSA, DMA, DGA, and Data Act. In every case, the impact of these initiatives on the media sector has been explored;

• The International AI Doctoral Academy (AIDA) has been successfully established with the support of its 53 founding members, including leading European universities, research centres and industry. Offering three distinct curricula (AI core, AI & Media, AI & Society) and a variety of AI courses, AIDA is already becoming a reference point on AI PhD-level education in Europe;

• Close collaboration with the AI4EU platform has already been established, aiming to enrich the platform with new AI4Media resources;

• The first AI4Media open call for equity-free funding of new research and applications for AI & Media has been launched on September 1st, 2021. 10 projects will be selected and will be funded with 50K euros each, after an evaluation process;

• We have successfully organised a series of theme-based public workshops that cover different aspects of AI, aiming to disseminate the project’s research outcomes but also to allow fruitful discussion among peers and experts on hot AI topics. In addition, we joined forces with ELISE to co-organise an AI Mellontology e-symposium in the context of AIDA to debate interesting AI society/industry/economy topics. All recordings are available on AI4Media’s and AIDA’s YouTube channels.

• During the next period we will focus on the development of the first version of the use case demonstrators and the integration of new AI features in them; the development of new AI technologies for media and society; the establishment of the European AI Media Observatory; the expansion of AIDA with new members, new AI courses and new educational activities; the mobility of young AI researchers in the context of our Junior Fellows Exchange Program; the implementation of the first 10 open call research & application projects and the launch of a 2nd Open Call!

Follow us on the rest of this 4-year journey to explore the opportunities offered by AI for the Media Sector and co-shape the European AI agenda!