MeMAD is a groundbreaking project that revolutionizes the way audiovisual content is managed and re-purposed. By providing novel methods for the efficient re-use of multilingual audiovisual content, MeMAD transforms video management and digital storytelling in broadcasting and media production. Unlike traditional automatic video description methods, MeMAD goes beyond by teaching machines to learn from humans.

This results in time-aligned semantic extraction of objects, as well as the recognition of action sequences and utilization of audio. The project aims to enhance the discoverability and findability of audiovisual data by developing innovative methods for accessing and utilizing the content. Additionally, MeMAD's technologies significantly reduce the costs of multimedia production, as tasks that previously required extensive hours of human labor can now be accomplished with just a few hours of machine work.

By interpreting the content and providing supporting links to other media assets, MeMAD makes audiovisual content more appealing and accessible to users. Using cutting-edge technologies including deep and recurrent neural networks, machine learning, artificial intelligence, and big data analysis, MeMAD develops advanced media description and linking techniques. By combining automatic efficiency with human accuracy, MeMAD enhances the process of digital storytelling and industrializes the re-purposing of existing media.

This visionary project received funding from the European Union's Horizon 2020 research and innovation program