Dr. Goëry Genty

Goëry Genty is a world-leading scientist in nonlinear and ultrafast optics.  His research focuses on the study of instabilities and nonlinear dynamics,broadband supercontinuum sources for imaging and sensing applications, as well as the design and development of novel ghost imaging techniques. He has published over 100 papers and won the IUPAP Young Scientist Prize in 2011. He is a fellow of the Optical Society of America.

Website – http://members.femto-st.fr/daniel-brunner/

Talk

Using Machine Learning to ‘Predict’ Extreme Events in Fiber-Optics Instabilities from Single-shot Spectral Measurements

Abstract

The study of instabilities that drive extreme events is central to nonlinear science. Perhaps, the most canonical form of nonlinear instabilities is modulation instability (MI) describing the exponential growth of a weak perturbation on top of a continuous background. In optical fibres, when driven initially by small-amplitude noise, MI has been shown to lead to the emergence of localized temporal breathers with random statistics. It has also been suggested that these dynamics may be associated with the emergence of extreme events or rogue waves. However, direct measurement in the time-domain of the breather properties is extremely challenging, requiring complex time-lens systems that typically suffer from drastic experimental constraints. Real-time spectral measurement techniques such as the dispersive Fourier transform (DFT) on the other hand are commonly used to measure ultrafast instabilities. Although relatively simple to implement, the DFT only provides spectral information. Here, we show how machine learning can overcome this restriction to study time-domain properties of optical fibre modulation instability based only on spectral intensity measurements. Specifically, we demonstrate that it is possible to train a supervised neural network to correlate the spectral and temporal properties of modulation instability using numerical simulations, and then apply the trained neural network to the analysis of high dynamic range experimental MI spectra and yield the temporal probability distribution for the highest peaks in the instability field.