DVPS

The Diversis Viis, Plurima Solvens (DVPS) project is focused on advancing Multimodal Foundation Models (MMFM)—AI systems capable of learning from and integrating multiple types of data.

OBJECTIVES

  • Develop generalizable methods that work across diverse modalities and application domains. 
  • Create a unified framework for MMFM development, incorporating formal characteristics of each modality 
  • Establish principled approaches for integrating new modalities into existing FM and expanding the multimodal capabilities of current models. 

VISION

  • Generates foundational knowledge about MMFM through rigorous experimentation and analysis.  
  • Delivers tested, generalizable methods for MMFM development, supported by open-source software libraries. 
  • Creates a series of algorithms and techniques that work across diverse application domains and facilitate the easy integration of new modalities into domain-specific FM. 
  • Considers a diverse environment for development, encompassing three substantially different core application domains: Cardiology, Geo-intelligence, and Language Communication. Furthermore, we introduce the concept of “surprise application domains” that will be added during the project, to force our methods to generalize beyond our initial assumptions, driving innovation in MMFM science. 

OUR ROLE

Our team focuses on federated learning for cardiology, enabling AI to learn from medical data while preserving patient privacy. This will help create more accurate, personalized, and secure AI-driven healthcare solutions. 

PARTNERS   

NO. INSTITUTION COUNTRY
1
Translated
Italy
2
The Alan Turing Institute
United Kingdom
3
University of Oxford
United Kingdom
4
Eidgenössische Technische Hochschule Zürich
Switserland
5
École Polytechnique Fédérale de Lausanne (EPFL)
Switserland
6
Karlsruhe Institute of Technology (KIT)
Germany
7
Fondazione Bruno Kessler (FBK)
Italy
8
Cyfronet HPC
Poland
9
Pi School
Italy
10
Heidelberg University Hospital
Germany
11
Meteorological and Environmental Earth Observation (MEEO)
Italy
12
Sistema GmbH
Austria
13
Universitat de Barcelona
Barcelona
14
Amsterdam UMC
Netherlands
15
Data Valley Consulting
Italy
16
Lynkeus
Italy
17
Vlaamse Instelling voor Technologisch Onderzoek NV (VITO)
Belgium
18
Vall d’Hebron Hospital Institute
Spain
19
Deepset
Germany
20
Imperial College London
United Kingdom
earlycause-logo
earlycause-logo

Funding institution

European Commision

Project number

101213369

Project budget

24,980,650 €

UB's budget

695,041.25 €

Project Coordinator

Translated SRL (Italy)

UB's Principal Investigator

Karim Lekadir

BCN-AIM's role

WP3 & WP7

Visit Our Website

https://dvps.translated.com/

X

https://x.com/dvpsai

Mail

karim.lekadir@ub.edu