About the Workshop

The Workshop on Physics-Informed Machine Learning (PIML) will be held 16–18 September 2026 at FORTH, Heraklion, Crete, Greece. It brings together leading researchers working at the intersection of machine learning, inverse problems, and physics-based modelling, with applications in Earth observation, medical imaging, and astrophysics.

Topics of interest include: physics-informed and hybrid AI models for inverse problems; deep unrolling and algorithm unfolding for signal reconstruction; neural operators and surrogate modelling for PDEs; generative models and learned priors for imaging and sensing; Bayesian inference and uncertainty quantification in physics-constrained learning; and applications in Earth observation, medical imaging, and astrophysics.

Participation does not require an abstract; researchers wishing to present may optionally submit an extended abstract (500 words max). Abstract deadline: 30 June 2026; notification: 5 July 2026; registration deadline: 10 September 2026.

Organizing Committee: J.-L. Starck, G. Tsagkatakis, P. Tsakalides, D. De Santis, L. G. Papale. The event is co-sponsored by the EU TITAN project and IEEE GRSS.