Embedded Files
Deep learning prediction of noise-driven nonlinear instabilities in fibre optics
Our latest work has just been published in Nature Communications ! In this work, we demonstrate that artificial neural networks can predict and bring a certain degree of control in incoherent dynamics—a longstanding challenge in nonlinear fibre optics.
By combining deep learning with optical seeding and real-time acquisition techniques, we manage to tune, predict and retrieve hidden information from noisy and fluctuating signals. A first step towards an extended control in shaping complex, noise-driven processes in photonics. A work done at XLIM (CNRS/ Université de Limoges), in collaboration with FEMTO-ST Institute and Leibniz Universität Hannover.
See the full article in Nature Communications (2025)
>> Full access to the training data: https://doi.org/10.5281/zenodo.15179897
New paper in ACS Photonics
Single-Photon Level Dispersive Fourier Transform: Ultrasensitive Characterization of Noise-Driven Nonlinear Dynamics
Our latest work with the team of Michael Kues (Leibniz University Hannover) is online, and featured in Cover of the ACS Photonics Issue of November 2023.
In this paper, we leverage quantum-inspired dispersive Fourier transform to characterize ultrafast spectral fluctuations with great sensitivity (< fW), high spectral resolution (~50 pm), and excellent dynamic range (up to ~80dB).
A great way to analyze and push forward the study of nonlinear instabilities and incoherent processes in photonics with improved performances.
See the full article in ACS Photonics (2023)
ERC Starting Grant 2020 awarded: STREAMLINE starts in February 2021 (for 5 years)
The project STREAMLINE was selected for funding by the ERC
Founded in 2007, the European Research Council (ERC) funds innovative exploratory projects that are likely to generate scientific, technological or societal progress. Evaluated by international experts, the ERC awards individual research grants to scientists based on scientific excellence as sole selection criterion.
See the STREAMLINE Poject Factsheet for more details
Page updated
Google Sites
Report abuse