LIP's participation in the AI4EOSC and iMagine projects
"AI4EOSC and iMagine will implement and explore artificial intelligence, deep learning and machine learning to better serve the scientific community."
Scientific and industrial research, as well as the growing use of massive amounts of data, justify the increasing investment in artificial intelligence, deep learning and machine learning in these fields. This is how the AI4EOSC (Artificial Intelligence for the European Open Science Cloud), and iMagine (Imaging Data and Services for Aquatic science) projects, which count with the participation of the LIP's distributed computing and digital infrastructures group, came about.
In the AI4EOSC project, LIP will contribute to the development, validation and availability of a platform for machine learning. It should be noted that LIP is fully responsible for software and quality assurance of services, applications and data. In the iMagine project, the same platform will be applied to use cases in the area of aquatic sciences. The iMagine project also counts with the participation of the National Infrastructure for Distributed Computing (INCD), in which LIP also participates in the development, operation and coordination.
Both projects started in September 2022. They are funded under the European Open Science Cloud (EOSC) to improve the supply of AI, deep learning and machine learning in Europe, providing the scientific community with access to these cutting-edge technologies.
One of the main tools that will contribute to this European push is the DEEP platform, initially developed in the DEEP-Hybrid-DataCloud project. Until now, the platform has been at the service of European researchers to enable them to develop and train machine learning and deep learning models.
The improvements already designed for the platform ensure new high-level functionalities, allowing users to save time to get results and increasing productivity by creating new analytical tools, products and services.
It should also be noted that one of the major commitments of this new project is to ensure that research products - data, models, metadata or publications - follow the principles of open science and FAIR (Findable, Accessible, Interoperable and Reusable) data, which are currently promoted within the EOSC.