Deep inverse photonic design: A tutorial

Yang Deng; Simiao Ren; Jordan Malof; Willie J. Padilla

doi:10.1016/j.photonics.2022.101070

Deep inverse photonic design: A tutorial

Yang Deng, Simiao Ren, Jordan Malof, Willie J. Padilla

Computer Science

Duke University

Research output: Contribution to journal › Article › peer-review

15 Scopus citations

Abstract

Deep learning (DL) has been used to design deep neural networks (DNNs) which have recently been applied to solving inverse problems in artificial electromagnetic materials (AEMs). Although inverse problems are often ill-posed, and therefore are difficult to solve, deep inverse models (DIMs) have achieved impressive results often surpassing capabilities possible with other approaches. We overview the process of deep inverse learning applied to AEM problems, including the building of data sets, design of a forward model, and comparison of inverse approaches including limitations. We conclude by detailing some important outstanding issues of deep inverse design of AEMs, and present an outlook of this exciting field.

Original language	English
Article number	101070
Journal	Photonics and Nanostructures - Fundamentals and Applications
Volume	52
DOIs	https://doi.org/10.1016/j.photonics.2022.101070
State	Published - Dec 2022

Funding

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Willie Padilla reports financial support was provided by US Department of Energy. We acknowledge support from the US Department of Energy ( DOE ) ( DESC0014372 ).

Funder number
DESC0014372

Keywords

Deep learning
Inverse
Photonics

Access to Document

10.1016/j.photonics.2022.101070

Cite this

@article{f7f7125cef584ec994e98409c1ad6531,

title = "Deep inverse photonic design: A tutorial",

abstract = "Deep learning (DL) has been used to design deep neural networks (DNNs) which have recently been applied to solving inverse problems in artificial electromagnetic materials (AEMs). Although inverse problems are often ill-posed, and therefore are difficult to solve, deep inverse models (DIMs) have achieved impressive results often surpassing capabilities possible with other approaches. We overview the process of deep inverse learning applied to AEM problems, including the building of data sets, design of a forward model, and comparison of inverse approaches including limitations. We conclude by detailing some important outstanding issues of deep inverse design of AEMs, and present an outlook of this exciting field.",

keywords = "Deep learning, Inverse, Photonics",

author = "Yang Deng and Simiao Ren and Jordan Malof and Padilla, \{Willie J.\}",

note = "Publisher Copyright: {\textcopyright} 2022",

year = "2022",

month = dec,

doi = "10.1016/j.photonics.2022.101070",

language = "English",

volume = "52",

journal = "Photonics and Nanostructures - Fundamentals and Applications",

issn = "1569-4410",

publisher = "Elsevier B.V.",

}

TY - JOUR

T1 - Deep inverse photonic design

T2 - A tutorial

AU - Deng, Yang

AU - Ren, Simiao

AU - Malof, Jordan

AU - Padilla, Willie J.

PY - 2022/12

Y1 - 2022/12

N2 - Deep learning (DL) has been used to design deep neural networks (DNNs) which have recently been applied to solving inverse problems in artificial electromagnetic materials (AEMs). Although inverse problems are often ill-posed, and therefore are difficult to solve, deep inverse models (DIMs) have achieved impressive results often surpassing capabilities possible with other approaches. We overview the process of deep inverse learning applied to AEM problems, including the building of data sets, design of a forward model, and comparison of inverse approaches including limitations. We conclude by detailing some important outstanding issues of deep inverse design of AEMs, and present an outlook of this exciting field.

AB - Deep learning (DL) has been used to design deep neural networks (DNNs) which have recently been applied to solving inverse problems in artificial electromagnetic materials (AEMs). Although inverse problems are often ill-posed, and therefore are difficult to solve, deep inverse models (DIMs) have achieved impressive results often surpassing capabilities possible with other approaches. We overview the process of deep inverse learning applied to AEM problems, including the building of data sets, design of a forward model, and comparison of inverse approaches including limitations. We conclude by detailing some important outstanding issues of deep inverse design of AEMs, and present an outlook of this exciting field.

KW - Deep learning

KW - Inverse

KW - Photonics

UR - https://www.scopus.com/pages/publications/85139361590

U2 - 10.1016/j.photonics.2022.101070

DO - 10.1016/j.photonics.2022.101070

M3 - Article

AN - SCOPUS:85139361590

SN - 1569-4410

VL - 52

JO - Photonics and Nanostructures - Fundamentals and Applications

JF - Photonics and Nanostructures - Fundamentals and Applications

M1 - 101070

ER -

Deep inverse photonic design: A tutorial

Abstract

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this