A look back at the first training session for “U-net image analysis“
Publié le
A look back at the first training session offered by the DIADEM Academy as part of the PEPR DIADEM continuing education program
Organized by Fabrice Rossignol and Elodie Iste, coordinators of the DIADEM Academy’s continuing education program, the training course entitled “U-net for image analysis” is part of a broader effort to disseminate the tools and skills needed to integrate digital approaches and artificial intelligence into accelerated materials discovery.
Our continuing education project aims to be a real lever for scientific exchange and cohesion, helping to build an interdisciplinary community that is both agile and collaborative.
In order to strengthen the link between training and the program’s research activities, we have decided to include presentations—and even visits when the location allows—of targeted DIADEM projects. The aim is to showcase the work carried out in the laboratories and to introduce participants to the digital tools currently under development, in particular the DIAMOND platform.
Presentation of targeted projects: ATHERM_COAT / DIAMS / DIAMOND
This time for discussion and discovery allowed participants to dive into the heart of the projects and gain a concrete understanding of how digital tools and AI approaches are transforming materials research.
They also helped to forge links between participants and research teams, while encouraging dialogue between disciplines and making technological advances and the tools developed tangible.
The DIAMOND platform, designed as part of the DIADEM PEPR, offers participants access to AI simulation workflows, structured databases, and containerized codes—tools that, once implemented, save considerable time in the discovery and development of new materials.
Focus on the “U-net for image analysis” training course
For this first session, participants had a variety of backgrounds—doctoral students, postdoctoral researchers, engineers, and researchers—bringing together a diversity of experiences and needs. They came from all over France to take part in this day dedicated to an introduction to U-net architecture and its applications in image analysis.
An expert trainer and educator
The training was led by Dr. Dmitry Karpov, Junior Professor of Physics and Materials Science at the University of Grenoble Alpes. A specialist in X-ray nanoimaging and computational imaging, he develops autonomous workflows combining digital models and machine learning to accelerate the analysis of experimental data and the discovery of new materials.
Highly appreciated for his professionalism and clear teaching style, he was able to make sometimes complex concepts accessible, while sharing concrete examples from his work. His structured approach, combining theoretical reminders, historical perspective, and guided exercises, allowed participants to progress at their own pace, regardless of their initial level. His willingness to answer questions and discuss each participant’s projects was also greatly appreciated.
Thank you to all participants and local teams.
We would particularly like to thank the PEPR DIADEM initial training team for their welcome and assistance: Sonia Collaud and Noël Jakse.
We would also like to thank Laurent Artaud (SIMAP laboratory) for presenting the ATHERM_COAT project, Digvijay Singh (IJL laboratory) for presenting DIAMS, and the entire DIAMOND team for their expertise, kindness, and availability: Irina Piazza, Joao Paulo Almeida de Mendoça, Léo Orveillon, and Martin Uhrin.
To stay informed about our news and upcoming events, sign up for our newsletter!
Plus d'actualités
