FAIR Workshop on Italian Visual and Language Models: Challenges and Activities
On February 5th, 2024, the Modena ELLIS Unit hosts the Workshop on Italian Visual and Language Models: challenges and activities, at AGO Modena Fabbriche Culturali (Piazzale Sant’Agostino, Modena).
This second workshop, after the one organized in Rome, focuses on the development of Large Multimodal Models (LMMs) for the Italian language, within the transversal project “Vision, Language, and Multimodal Challenges” of the FAIR – Future Artificial Intelligence Research project. The goal is to discuss with the Italian community the future challenges in the design, development and evaluation of Large Language Models, Large Visual Models and their combination, and their application to domains such as health and manufacturing. Some results of FAIR research activities are also presented.
Organizers
Prof. Rita Cucchiara (CNR – UNIMORE) and Prof. Roberto Navigli (Sapienza University of Rome).
Programme
| 09:10 – 09:30 | WelcomeStefano Bonaccini (President of the Emilia-Romagna Region), Carlo Adolfo Porro (Rector of UNIMORE), Donatella Pieri (President of Fondazione AGO) |
| 09:30 – 09:45 | The project and the results of the Rome workshopRoberto Navigli |
| 09:45 – 12:00 | Challenges and activities on Large Visual and Language Models in FAIRRita Cucchiara, Roberto Navigli and the FAIR partners |
| 12:00 – 12:30 | The present and future of the collaboration between FAIR and CINECASanzio Bassini (CINECA) |
| 12:30 – 13:00 | AI e il Visconte CiberneticoMassimo Sideri (Corriere della Sera) |
| 13:00 – 14:30 | Lunch and poster presentations |
| 14:30 – 15:00 | Open foundation models: reproducible science of transferable learningJenia Jitsev and Marianna Nezhurina (LAION) |
| 15:00 – 15:30 | FAIR results on LLMs and LMMsGiovanni Maria Farinella (visual datasets), Lorenzo Baraldi (the MORE Italian vision-and-language model), Pierpaolo Basile and Giovanni Semeraro (LLaMAntino) |
| 15:30 – 17:30 | Open discussion and conclusionsMultimodal adapters and foundation models, application domains and datasets, evaluation and trustworthiness, a FAIR position paper on multimodal foundation models |
The workshop is supported by the National PhD Program in “AI for Society”, Modena site.