A physics-defined recurrent neural network to compute coherent light wave scattering on the millimetre scale
arXiv:2208.01118 · doi:10.34133/icomputing.0098
Abstract
Heterogeneous materials such as biological tissue scatter light in random, yet deterministic, ways. Wavefront shaping can reverse the effects of scattering to enable deep-tissue microscopy. Such methods require either invasive access to the internal field or the ability to numerically compute it. However, calculating the coherent field on a scale relevant to microscopy remains excessively demanding for consumer hardware. Here we show how a recurrent neural network can mirror Maxwell's equations without training. By harnessing public machine learning infrastructure, such \emph{Scattering Network} can compute the -wavelength light field throughout a or scattering volume. The elimination of the training phase cuts the calculation time to a minimum and, importantly, it ensures a fully deterministic solution, free of any training bias. The integration with an open-source electromagnetic solver enables any researcher with an internet connection to calculate complex light-scattering in volumes that are larger by two orders of magnitude
arXiv comment8 pages, 3 figures
References in corpus (23)
- PyTorch: An Imperative Style, High-Performance Deep Learning Library
- Measuring the Transmission Matrix in Optics : An Approach to the Study and Control of Light Propagation in Disordered Media
- Outlook for inverse design in nanophotonics
- Exploiting disorder for perfect focusing
- Light fields in complex media: mesoscopic scattering meets wave control
- Wave Physics as an Analog Recurrent Neural Network
- Physically-informed artificial neural networks for atomistic modeling of materials
- The recursive Green's function method for graphene
- Roadmap on Wavefront Shaping and deep imaging in complex media
- Quantum Simulation of Coherent Hawking-Unruh Radiation
- A convergent Born series for solving the inhomogeneous Helmholtz equation in arbitrarily large media
- Anderson localization of electromagnetic waves in three dimensions
- Depth-Targeted Energy Deposition Deep Inside Scattering Media
- Tracking moving objects through scattering media via speckle correlations
- Fast multi-source nanophotonic simulations using augmented partial factorization
- How to build the optical inverse of a multimode fibre
- Coherent enhancement of optical remission in diffusive media
- Calculating coherent light-wave propagation in large heterogeneous media
- MaxwellNet: Physics-driven deep neural network training based on Maxwell's equations
- Energy-leaky modes in partially measured scattering matrices of disordered media
- A physics-defined recurrent neural network to compute coherent light wave scattering on the millimetre scale
- A universal matrix-free split preconditioner for the fixed-point iterative solution of non-symmetric linear systems
- Large area optimization of meta-lens via data-free machine learning