From May 14 to 16, 2025, the first SensAIfood Training School took place in Namur and Gembloux. This training course focused on artificial intelligence methods applied to spectral data analysis. Organized as part of the COST Action IG19145 – SensAIfood (Innovative spectral sensing and AI technologies for food quality, safety and authenticity), the event aimed to strengthen researchers’ skills in a key area for innovation in the agri-food sector.
This first edition, coordinated by the DRIM laboratory of U12 (quality and authentication of agricultural products), brought together 33 participants from various countries, including researchers, PhD students, lecturers, and domain experts. The three-day program, split between Namur and Gembloux, offered a balanced mix of theoretical content, applications, and hands-on sessions.
The main topics covered during the training included:
- An introduction to artificial intelligence methods applied to spectral data,
- the differences between supervised and unsupervised learning,
- the implementation of artificial neural networks (MLP, CNN),
- the application of deep learning to real-world spectral analysis cases,
- critical insights into best practices for deploying these methods,
- a guided practical session allowing participants to work with real data using appropriate tools.
Beyond providing a strong foundation of knowledge, the training fostered exchanges across disciplines (spectroscopy, chemometrics, AI) and sectors (academic research and industrial applications).
This first edition, led by CRA-W as part of the COST Action IG19145, represents an important step in spreading skills and best practices in artificial intelligence applied to spectroscopy. It illustrates the center’s commitment to international scientific collaboration.
