Probabilistic inverse design of metasurfaces using mixture density neural networks

Metasurfaces are planar sub-micron structures that can outperform traditional optical elements and miniaturize optical devices. Optimization-based inverse designs of metasurfaces often get trapped in a local minimum, and the inherent non-uniqueness property of the inverse problem plagues approaches...

Full description

Saved in:
Bibliographic Details
Main Authors: Mahsa Torfeh, Chia Wei Hsu
Format: Article
Language:English
Published: IOP Publishing 2024-01-01
Series:JPhys Photonics
Subjects:
Online Access:https://doi.org/10.1088/2515-7647/ad9b82
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850114406837387264
author Mahsa Torfeh
Chia Wei Hsu
author_facet Mahsa Torfeh
Chia Wei Hsu
author_sort Mahsa Torfeh
collection DOAJ
description Metasurfaces are planar sub-micron structures that can outperform traditional optical elements and miniaturize optical devices. Optimization-based inverse designs of metasurfaces often get trapped in a local minimum, and the inherent non-uniqueness property of the inverse problem plagues approaches based on conventional neural networks. Here, we use mixture density neural networks to overcome the non-uniqueness issue for the design of metasurfaces. Once trained, the mixture density network (MDN) can predict a probability distribution of different optimal structures given any desired property as the input, without resorting to an iterative local optimization. As an example, we use the MDN to design metasurfaces that project structured light patterns with varying fields of view. This approach enables an efficient and reliable inverse design of fabrication-ready metasurfaces with complex functionalities without getting trapped in local optima.
format Article
id doaj-art-29d27d2ecb7842079bbcf927579e7a7f
institution OA Journals
issn 2515-7647
language English
publishDate 2024-01-01
publisher IOP Publishing
record_format Article
series JPhys Photonics
spelling doaj-art-29d27d2ecb7842079bbcf927579e7a7f2025-08-20T02:36:53ZengIOP PublishingJPhys Photonics2515-76472024-01-017101500710.1088/2515-7647/ad9b82Probabilistic inverse design of metasurfaces using mixture density neural networksMahsa Torfeh0https://orcid.org/0009-0003-6782-9331Chia Wei Hsu1Ming Hsieh Department of Electrical and Computer Engineering, University of Southern California , Los Angeles, CA 90089, United States of AmericaMing Hsieh Department of Electrical and Computer Engineering, University of Southern California , Los Angeles, CA 90089, United States of AmericaMetasurfaces are planar sub-micron structures that can outperform traditional optical elements and miniaturize optical devices. Optimization-based inverse designs of metasurfaces often get trapped in a local minimum, and the inherent non-uniqueness property of the inverse problem plagues approaches based on conventional neural networks. Here, we use mixture density neural networks to overcome the non-uniqueness issue for the design of metasurfaces. Once trained, the mixture density network (MDN) can predict a probability distribution of different optimal structures given any desired property as the input, without resorting to an iterative local optimization. As an example, we use the MDN to design metasurfaces that project structured light patterns with varying fields of view. This approach enables an efficient and reliable inverse design of fabrication-ready metasurfaces with complex functionalities without getting trapped in local optima.https://doi.org/10.1088/2515-7647/ad9b82Nanophotonicsmetasurfaceinverse designdeep neural networkmixture density networkstructured light
spellingShingle Mahsa Torfeh
Chia Wei Hsu
Probabilistic inverse design of metasurfaces using mixture density neural networks
JPhys Photonics
Nanophotonics
metasurface
inverse design
deep neural network
mixture density network
structured light
title Probabilistic inverse design of metasurfaces using mixture density neural networks
title_full Probabilistic inverse design of metasurfaces using mixture density neural networks
title_fullStr Probabilistic inverse design of metasurfaces using mixture density neural networks
title_full_unstemmed Probabilistic inverse design of metasurfaces using mixture density neural networks
title_short Probabilistic inverse design of metasurfaces using mixture density neural networks
title_sort probabilistic inverse design of metasurfaces using mixture density neural networks
topic Nanophotonics
metasurface
inverse design
deep neural network
mixture density network
structured light
url https://doi.org/10.1088/2515-7647/ad9b82
work_keys_str_mv AT mahsatorfeh probabilisticinversedesignofmetasurfacesusingmixturedensityneuralnetworks
AT chiaweihsu probabilisticinversedesignofmetasurfacesusingmixturedensityneuralnetworks