Generative model for multiple-purpose inverse design and forward prediction of disordered waveguides in linear and nonlinear regimes

Ziheng Guo; Zhongliang Guo; Oggie Arandelovic; Andrea di Falco

doi:10.1117/12.3017130

Generative model for multiple-purpose inverse design and forward prediction of disordered waveguides in linear and nonlinear regimes

Ziheng Guo, Zhongliang Guo, Oggie Arandelovic, Andrea di Falco^*

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Data-driven machine learning framework has become a state-of-art technique to explore whole parameters design space for designing complex systems. In this work, we used conditional generative adversarial networks to inverse design three problems that we are interested in random nanophotonic systems: pattern optimization, geometry generation, and pattern reproduction. Meanwhile, automation convolutional neural networks group for forward prediction of the transmission spectra of disordered waveguides in linear and nonlinear regimes, at telecommunication wavelengths.

Original language	English
Title of host publication	Machine Learning in Photonics
Editors	Francesco Ferranti, Mehdi Keshavarz Hedayati, Andrea Fratalocchi
Place of Publication	Bellingham, WA
Publisher	SPIE
Number of pages	5
ISBN (Electronic)	9781510673533
ISBN (Print)	9781510673526
DOIs	https://doi.org/10.1117/12.3017130
Publication status	Published - 18 Jun 2024
Event	SPIE Photonics Europe 2024 - Strasbourg, France Duration: 7 Apr 2024 → 12 Apr 2024

Publication series

Name	Proceedings of SPIE - The International Society for Optical Engineering
Volume	13017
ISSN (Print)	0277-786X
ISSN (Electronic)	1996-756X

Conference

Conference	SPIE Photonics Europe 2024
Country/Territory	France
City	Strasbourg
Period	7/04/24 → 12/04/24

Keywords

Design
Waveguides
Network architectures
Neural networks
Gallium nitride
Optical transmission
Complex systems
Finite-difference time-domain method
Nonlinear transmission
Systems modeling

Access to Document

10.1117/12.3017130

Guo_2024_MLiP_Generative-model-inverse-design-prediction-disordered-waveguides_AAM_CCBY
Copyright © 2024 the Authors. This work has been made available online in accordance with the University of St Andrews Open Access policy. This accepted manuscript is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. The final published version of this work is available at https://doi.org/10.1117/12.3017130
Accepted author manuscript, 657 KBLicence: CC BY

Handle.net

Persistent link

Cite this

Guo, Z., Guo, Z., Arandelovic, O., & di Falco, A. (2024). Generative model for multiple-purpose inverse design and forward prediction of disordered waveguides in linear and nonlinear regimes. In F. Ferranti, M. K. Hedayati, & A. Fratalocchi (Eds.), Machine Learning in Photonics Article 1301702 (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 13017). SPIE. https://doi.org/10.1117/12.3017130

Guo, Ziheng ; Guo, Zhongliang ; Arandelovic, Oggie et al. / Generative model for multiple-purpose inverse design and forward prediction of disordered waveguides in linear and nonlinear regimes. Machine Learning in Photonics. editor / Francesco Ferranti ; Mehdi Keshavarz Hedayati ; Andrea Fratalocchi. Bellingham, WA : SPIE, 2024. (Proceedings of SPIE - The International Society for Optical Engineering).

@inproceedings{652618cbd2734e7d8beb556c59cbd4b4,

title = "Generative model for multiple-purpose inverse design and forward prediction of disordered waveguides in linear and nonlinear regimes",

abstract = "Data-driven machine learning framework has become a state-of-art technique to explore whole parameters design space for designing complex systems. In this work, we used conditional generative adversarial networks to inverse design three problems that we are interested in random nanophotonic systems: pattern optimization, geometry generation, and pattern reproduction. Meanwhile, automation convolutional neural networks group for forward prediction of the transmission spectra of disordered waveguides in linear and nonlinear regimes, at telecommunication wavelengths.",

keywords = "Design, Waveguides, Network architectures, Neural networks, Gallium nitride, Optical transmission, Complex systems, Finite-difference time-domain method, Nonlinear transmission, Systems modeling",

author = "Ziheng Guo and Zhongliang Guo and Oggie Arandelovic and \{di Falco\}, Andrea",

year = "2024",

month = jun,

day = "18",

doi = "10.1117/12.3017130",

language = "English",

isbn = "9781510673526",

series = "Proceedings of SPIE - The International Society for Optical Engineering",

publisher = "SPIE",

editor = "Francesco Ferranti and Hedayati, \{Mehdi Keshavarz\} and Andrea Fratalocchi",

booktitle = "Machine Learning in Photonics",

address = "United States",

note = "SPIE Photonics Europe 2024 ; Conference date: 07-04-2024 Through 12-04-2024",

}

Guo, Z, Guo, Z, Arandelovic, O & di Falco, A 2024, Generative model for multiple-purpose inverse design and forward prediction of disordered waveguides in linear and nonlinear regimes. in F Ferranti, MK Hedayati & A Fratalocchi (eds), Machine Learning in Photonics., 1301702, Proceedings of SPIE - The International Society for Optical Engineering, vol. 13017, SPIE, Bellingham, WA, SPIE Photonics Europe 2024, Strasbourg, France, 7/04/24. https://doi.org/10.1117/12.3017130

Generative model for multiple-purpose inverse design and forward prediction of disordered waveguides in linear and nonlinear regimes. / Guo, Ziheng; Guo, Zhongliang; Arandelovic, Oggie et al.
Machine Learning in Photonics. ed. / Francesco Ferranti; Mehdi Keshavarz Hedayati; Andrea Fratalocchi. Bellingham, WA: SPIE, 2024. 1301702 (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 13017).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Generative model for multiple-purpose inverse design and forward prediction of disordered waveguides in linear and nonlinear regimes

AU - Guo, Ziheng

AU - Guo, Zhongliang

AU - Arandelovic, Oggie

AU - di Falco, Andrea

PY - 2024/6/18

Y1 - 2024/6/18

N2 - Data-driven machine learning framework has become a state-of-art technique to explore whole parameters design space for designing complex systems. In this work, we used conditional generative adversarial networks to inverse design three problems that we are interested in random nanophotonic systems: pattern optimization, geometry generation, and pattern reproduction. Meanwhile, automation convolutional neural networks group for forward prediction of the transmission spectra of disordered waveguides in linear and nonlinear regimes, at telecommunication wavelengths.

AB - Data-driven machine learning framework has become a state-of-art technique to explore whole parameters design space for designing complex systems. In this work, we used conditional generative adversarial networks to inverse design three problems that we are interested in random nanophotonic systems: pattern optimization, geometry generation, and pattern reproduction. Meanwhile, automation convolutional neural networks group for forward prediction of the transmission spectra of disordered waveguides in linear and nonlinear regimes, at telecommunication wavelengths.

KW - Design

KW - Waveguides

KW - Network architectures

KW - Neural networks

KW - Gallium nitride

KW - Optical transmission

KW - Complex systems

KW - Finite-difference time-domain method

KW - Nonlinear transmission

KW - Systems modeling

UR - https://www.scopus.com/pages/publications/85200210913

U2 - 10.1117/12.3017130

DO - 10.1117/12.3017130

M3 - Conference contribution

AN - SCOPUS:85200210913

SN - 9781510673526

T3 - Proceedings of SPIE - The International Society for Optical Engineering

BT - Machine Learning in Photonics

A2 - Ferranti, Francesco

A2 - Hedayati, Mehdi Keshavarz

A2 - Fratalocchi, Andrea

PB - SPIE

CY - Bellingham, WA

T2 - SPIE Photonics Europe 2024

Y2 - 7 April 2024 through 12 April 2024

ER -

Guo Z, Guo Z, Arandelovic O , di Falco A. Generative model for multiple-purpose inverse design and forward prediction of disordered waveguides in linear and nonlinear regimes. In Ferranti F, Hedayati MK, Fratalocchi A, editors, Machine Learning in Photonics. Bellingham, WA: SPIE. 2024. 1301702. (Proceedings of SPIE - The International Society for Optical Engineering). doi: 10.1117/12.3017130

Generative model for multiple-purpose inverse design and forward prediction of disordered waveguides in linear and nonlinear regimes

Abstract

Publication series

Conference

Keywords

Access to Document

Handle.net

Other files and links

Fingerprint

Building trustworthy computer vision: adversarial techniques for robustness assessment and misuse prevention

Machine learning driven design and fabrication of photonic devices with engineered dispersion

Cite this

Generative model for multiple-purpose inverse design and forward prediction of disordered waveguides in linear and nonlinear regimes

Abstract

Publication series

Conference

Keywords

Access to Document

Handle.net

Other files and links

Fingerprint

Student theses

Building trustworthy computer vision: adversarial techniques for robustness assessment and misuse prevention

Machine learning driven design and fabrication of photonic devices with engineered dispersion

Cite this