International experts gather in Barcelona to discuss the latest advances in multimodal AI foundation models

Multimodal Foundation Models represent one of the most advanced research frontiers in artificial intelligence. These technologies can learn from vast amounts of heterogeneous data —including text, images, audio, video, and sensor data— enabling them to integrate and relate information from multiple sources to create systems that are more flexible, robust and closer to how humans perceive and understand the world. Their development is accelerating innovation across sectors such as healthcare, education, media, mobility, and intelligent assistants.

In the current European landscape, multimodal foundation models have become a strategic priority in the drive to develop competitive, open and sovereign artificial intelligence. This effort is supported by a coordinated ecosystem of research initiatives, including ELLIS, ELIAS, and ELLIOT. ELLIS serves as a scientific network connecting leading researchers in machine learning and AI across Europe; ELIAS, funded by the European Union, brings together research institutions to promote the transfer of scientific advances into applications with sustainable societal impact; and ELLIOT is a European project focused on developing open, trustworthy multimodal foundation models aligned with European values.

Within this framework, research in multimodal artificial intelligence seeks not only to advance technical capabilities but also to ensure that these breakthroughs translate into applied innovation, economic impact, and tangible societal benefits for Europe, with a strong emphasis on robustness, transparency, and responsibility.

Against this backdrop, Barcelona is emerging as one of Europe's leading hubs for artificial intelligence research. The city has established itself as a strategic node connecting universities, research centres, and cutting-edge European initiatives, fostering collaboration between science, technology, and industry. Barcelona is home to ELLIS Barcelona, which brings together leading AI researchers based in Catalonia and connects them with the wider ELLIS network. In addition, in 2023, with the support of the Government of Catalonia, the RDI-IA Network was launched as part of the Catalonia.AI strategy to promote innovation, technology transfer, and knowledge valorisation in artificial intelligence across the region. The network has helped strengthen a previously fragmented AI ecosystem while reinforcing links with Europe's leading research and innovation initiatives, consolidating Barcelona's position within the ELIAS Alliance.

Within this context, Barcelona hosted the workshop "Multimodal Foundation Models: From Research to Innovation" an international event that brought together researchers and industry representatives to discuss the latest advances in multimodal foundation models and their potential to transform scientific research into innovations with real impact on society and industry.

The programme featured talks from internationally recognised experts, including Dr Ioannis Patras (Queen Mary University of London), Dr Dima Damen (University of Bristol), Dr Joost van de Weijer (Computer Vision Center), Dr Jai Bardhan (Czech Institute of Informatics, Robotics and Cybernetics), and Dr Vittorio Murino (University of Verona). The industrial perspective was represented by Èric Verdaguer (LogMeal), who showcased real-world applications of multimodal AI in business environments.

The event also included a panel discussion addressing the challenges of translating cutting-edge research into innovative multimodal AI solutions. The panel brought together Jordina Torrents (HP), Lukas Fischer (NXAI), Isabelle Siegrist (Sandborn), and Èric Verdaguer (LogMeal), and was moderated by Dimosthenis Karatzas, Associate Director of the Computer Vision Center and Co-Director of ELLIS Barcelona.

The workshop opened with institutional remarks by Javier Selva, Director General for Technological Innovation and Scientific Internationalisation at the Department of Research and Universities of the Government of Catalonia. Closing remarks were delivered by Rafael González, Head of the Knowledge Transfer Promotion Area within the same department.

The event also featured a presentation of the RDI-IA Network by its Scientific Director, Dimosthenis Karatzas, highlighting the network's initiatives to foster knowledge transfer, technology valorisation, and collaboration between researchers, research groups, and industry.

The workshop welcomed nearly 150 participants to the Prat de la Riba Hall at the Institute for Catalan Studies, reaching full capacity, while around 100 additional attendees followed the event via live streaming.

The workshop was organised by the Computer Vision Center (CVC) in collaboration with ELLIS Barcelona, the ELLIS Multimodal Learning Systems Programme, the European projects ELLIOT and ELIAS, and the RDI-IA Network, with the support of Barcelona City Council.