
SAR Denoiser
A deep learning model that cleans a single Sentinel-1 radar image in one pass, trained on real multi-temporal references instead of synthetic noise. Presented as first author at ESA INSIGHT2026.
Starion Group · 2025 – 2026
Salvatore Prochilo · AI Engineer · Rome, Italy
Restoration models for Sentinel-1 radar and telescope imagery at Starion Group, after three years shipping production software. First author of an oral talk at ESA INSIGHT2026.
Case studies

A deep learning model that cleans a single Sentinel-1 radar image in one pass, trained on real multi-temporal references instead of synthetic noise. Presented as first author at ESA INSIGHT2026.
Starion Group · 2025 – 2026

MSc thesis: a fully convolutional network that uses Fast Fourier Convolutions to separate ink from paper in degraded historical documents, evaluated on ten DIBCO benchmarks.
UNIMORE – AImageLab · 2023
Talks & publications
First author, oral presentation · INSIGHT2026 – Innovative Signal Exploitation for Geospatial Intelligence
European Space Agency – ESRIN, Centre for Earth Observation · Frascati, Italy · 05/2026
10/2024 – present
Starion Group
Frascati (Rome), Italy
08/2021 – 09/2024
Capgemini Engineering
Rome, Italy
2023
Università degli Studi di Modena e Reggio Emilia
Modena, Italy
2017
Università della Calabria
Cosenza, Italy