Siwar Gharbi (PhD student at Xlim Research Institute, France – EMIMEP alumni) delivered a talk entilted « Smart Light Transport Through Hollow-Core Multimode Fibers Using Deep Learning » on October 23rd, 2025.

Bio. Siwar GHARBI is a PhD candidate in optics and photonics under a CIFRE industrial fellowship between Alphanov and the XLIM Laboratory, within the Phocal research group at the University of Limoges.

Her research explores AI-assisted control of structured light in hollow-core multimode fibres, with applications in high-power laser delivery and adaptive beam shaping.

Abstract. This PhD project aims to develop machine learning approaches for the shaping and dynamic control of light beams guided through hollow-core multimode optical fibres (HC-MMFs). By training neural networks to model the complex light transmission within these fibres, the method enables the prediction and generation of arbitrary output patterns from a spatially modulated input beam. This ongoing work contributes to the development of intelligent photonic systems capable of exciting and reconfiguring structured light with high precision at the fibre output, for applications in optical power delivery, and advanced laser technologies.