Optimization, neural networks and inverse design: from nanophotonics to computational imaging

by Prof. Giovanni Pellegrini

Friday 28 Mar 2025, 11:00 → 12:00 Europe/Rome

1/1-2 - Aula "C. Voci" (Dipartimento di Fisica e Astronomia - Edificio Marzolo)

The seminar will explore the use of optimization, neural networks, and inverse design in nanophotonics, as well as the role of neural networks in computational imaging. In nanophotonics, optimization and inverse design techniques enable the creation of structures with tailored optical properties, while deep learning provides new approaches for design and analysis. The same machine learning techniques also play a growing role in computational imaging, where they help reconstruct images and extract information beyond conventional limits. We will discuss specific examples, including the design of one-dimensional photonic crystals, imaging through complex media, and ultrafast imaging. The talk will cover methodologies, recent advancements, and practical applications of these approaches.

Seminario_Pellegrini.pdf