Month: August 2024

Artificial Intelligence (AI) is revolutionising the media industry, enhancing content creation, analysis, and distribution, and providing deeper insights into complex social dynamics both online and offline. AI empowers the media industry to support a democratic society, enrich our lives, and foster creativity like never before. However, the deployment of AI also introduces significant ethical challenges and risks for individuals and society at large.

In response to these opportunities and challenges, the AI4Media Network of Excellence has established an extensive network of researchers across Europe and beyond, dedicated to exploring AI’s applications in media, society, and democracy. Over nearly four years, AI4Media has become a beacon of innovation, collaboration, and progress, linking AI with the media industry.

AI4Media’s Four Pillars of Success:

1. A Dynamic Network (LINK): Comprising over 200 organisations from academia, industry, and civil society, this network fosters a collaborative environment for sharing ideas, resources, and new developments. Associate members benefit from joint projects, a vast pool of AI knowledge, and enhanced visibility through cross-organizational exchanges and access to educational materials.
2. AI Media Observatory (LINK): Serving as a crucial resource for stakeholders, the Observatory offers up-to-date information and insights on AI in media. It identifies trends and challenges, supporting informed decision-making and fostering responsible AI development. The Observatory is useful for media professionals, AI researchers, policymakers, educators, and anyone interested in AI’s impact on media.
3. International AI Doctoral Academy (AIDA) (LINK): Enhancing AI education across Europe, AIDA attracts PhD students and promotes academic excellence. It provides a variety of educational programs and resources, standardising AI education across European universities and supporting collaborative opportunities. AIDA has offered over 80 educational courses attended by more than 2,000 students/researchers/professionals and provides a repository with over 300 educational resources.
4. Scientific and Research Outputs (LINK): AI4Media has made substantial contributions through publications, open datasets, software tools, and strategic documents. These resources drive innovation in AI and are vital for researchers, AI professionals, media stakeholders, policymakers, and educators. AI4Media’s research activities have yielded over 400 scientific publications, numerous reports, nearly 100 open-access software tools, and more than 40 open datasets. The project communicates its key outputs through factsheets, booklets, and infographics, collectively referred to as “results in brief.”

AI4Media’s Key Contributions

• AI4Media Strategic Research Agenda (SRA) (LINK): The AI4Media SRA outlines strategic planning for AI research and community engagement. Initially released in March 2023 and updated in July 2024, the SRA includes new developments in Generative AI, driven by the widespread use of Large Language Models and Large Multimodal Models. It addresses significant opportunities and challenges these technologies present to media, creative industries, and society.

• AI4Media Roadmap on AI Technologies and Applications for the Media Industry (LINK): The AI4Media Roadmap explores how AI technologies are transforming the media sector and provides insights into future developments. It examines opportunities for growth and transformation while addressing relevant risks and mitigation strategies.

• AI4Media Junior Fellows Exchange Program (LINK): The program facilitates exchanges of early career researchers to improve their skills and knowledge in AI for the media and society. It has facilitated 88 exchanges involving 65 organisations across Europe and beyond, producing important outcomes in publications, open software, and datasets.

• AI4Media Open Call #1 (LINK) and Open Call #2 (LINK): AI4Media launched two Open Calls targeting entrepreneurs, companies, and researchers. From 155 applications, 20 projects were funded, providing up to €50,000 in equity-free funding, coaching, and promotional services. These projects developed innovative methods and applications addressing significant challenges in media and AI.

• AI4Media Use Cases (LINK): AI4Media implemented 7 use cases in collaboration with European media organisations. These use cases addressed disinformation, news production, media moderation, audiovisual archives, game design, artistic co-creation, and social science research. They highlight AI’s application across the media industry value chain.

• AI4Media Technological Highlights: Four booklets showcase AI4Media’s key technological highlights:

• • Human- and Society-Centered AI – (LINK)
  • New Learning Paradigms & Distributed AI – (LINK)
  • Trustworthy AI – (LINK)
  • Content-Centered AI – (LINK)

• Video Series “AI Applications for the Media Industry” (LINK): AI4Media produced a series of videos covering AI applications in news production, robot journalism, social media, entertainment/movie production, games, music, and publishing.

• LLMaker: Consistent Game Content Creation (LINK): LLMaker is a specialised tool for iterative game content co-design through chat interactions. It bridges human designers and content generation systems, leveraging large language models to interpret designer requests into actionable instructions.

Through these initiatives, AI4Media has left a lasting legacy in the AI and media landscape, providing a wealth of resources, research, and educational opportunities. These efforts ensure that AI continues to advance responsibly and beneficially for all segments of society.

AI4Media is excited to announce the release of a series of booklets featuring the key technological and research highlights from the AI4Media project. These four comprehensive booklets aim to advance the next generation of core AI research for the media industry and society, enhancing Europe’s capabilities in media AI. Each booklet focuses on significant, groundbreaking, or impactful developments across different core AI research areas developed under AI4Media.

1. Booklet on Human- and Society-Centered AI – [LINK]

This booklet explores AI4Media’s contributions toward understanding the broad societal impacts and individual effects of AI. As AI technologies become increasingly prevalent in everyday applications, these insights are crucial for ensuring responsible deployment. The booklet delves into how AI can be harnessed to benefit society while addressing ethical considerations and potential risks.

2. Booklet on New Learning Paradigms & Distributed AI – [LINK]

This booklet presents innovative learning paradigms specifically tailored for media content and applications. These advancements push the boundaries of deep learning and state-of-the-art AI technologies. It highlights novel methodologies and frameworks that enhance the efficiency and effectiveness of AI systems in processing and understanding media content.

3. Booklet on Trustworthy AI – [LINK]

This booklet details new techniques and algorithms developed by AI4Media to foster trustworthy AI for media industry applications. It covers four key areas: Adversarial Robustness, Explainability and Interpretability, Privacy and Security, and Fairness. Additionally, this booklet introduces a novel AI benchmarking tool developed as part of AI4Media, which sets new standards for evaluating AI systems’ reliability and transparency.

4. Booklet on Content-Centered AI – [LINK]

This booklet presents key outcomes on the AI’s role in analysing, generating, and managing media content. It emphasizes technologies that enhance content creation and dissemination in the media sector. The insights provided in this booklet showcase how AI can revolutionize media workflows, from content production to audience engagement.

Each booklet serves as a valuable resource for stakeholders in the AI and media industries, providing in-depth insights into AI4Media’s leading-edge research and its potential applications. These booklets are designed to inform and inspire researchers, industry professionals, policymakers, and educators about the transformative potential of AI in the media sector.

By sharing these advancements, AI4Media continues to contribute to the development of responsible and impactful AI technologies that support the media industry and broader societal goals. We invite you to explore these booklets and discover the innovative work being done to shape the future of AI and media.