Predicting nonlinear optical scattering with physics-driven neural networks

TL;DR

This paper introduces a tunable physics-driven neural network, MaxwellNet, capable of rapidly predicting nonlinear optical scattering in media with Kerr effects by dynamically adjusting its weights based on intensity-dependent refractive indices.

Contribution

The authors develop a dynamic, tunable version of MaxwellNet that models nonlinear optical scattering with Kerr effects, enabling fast and accurate predictions for inhomogeneous media.

Findings

MaxwellNet can predict light scattering in nonlinear media within milliseconds.

The network dynamically adjusts to intensity-dependent refractive indices.

It effectively models Kerr effect-induced nonlinear optical phenomena.

Abstract

Deep neural networks trained on physical losses are emerging as promising surrogates of nonlinear numerical solvers. These tools can predict solutions of Maxwell's equations and compute gradients of output fields with respect to the material and geometrical properties in millisecond times which makes them attractive for inverse design or inverse scattering applications. Here we develop a tunable version of MaxwellNet, a physics driven neural network able to compute light scattering from inhomogenous media with a size comparable with the incident wavelength in the presence of the optical Kerr effect. The weights of the network are dynamically adjusted to take into account the intensity-dependent refractive index of the material.