Inverse deep learning methods and benchmarks for artificial electromagnetic material design
Abstract
In this work we investigate the use of deep inverse models (DIMs) for designing artificial electromagnetic materials (AEMs) – such as metamaterials, photonic crystals, and plasmonics – to achieve some desired scattering properties (e.g., transmission or reflection spectrum). DIMs are deep neural networks (i.e., deep learning models) that are specially-designed to solve ill-posed inverse problems. There has recently been tremendous growth in the use of DIMs for solving AEM design problems however there has been little comparison of these approaches to examine their absolute and relative performance capabilities. In this work we compare eight state-of-the-art DIMs on three unique AEM design problems, including two models that are novel to the AEM community. Our results indicate that DIMs can rapidly produce accurate designs to achieve a custom desired scattering on all three problems. Although no single model always performs best, the Neural-Adjoint approach achieves the best overall performance across all problem settings. As a final contribution we show that not all AEM design problems are ill-posed, and in such cases a conventional deep neural network can perform better than DIMs. We recommend that a deep neural network is always employed as a simple baseline approach when addressing AEM design problems. Furthermore,more »
- Authors:
-
- Duke University, Durham, NC (United States)
- Publication Date:
- Research Org.:
- Duke Univ., Durham, NC (United States)
- Sponsoring Org.:
- USDOE Office of Science (SC)
- OSTI Identifier:
- 1978815
- Alternate Identifier(s):
- OSTI ID: 1846707
- Grant/Contract Number:
- SC0014372; DESC0014372
- Resource Type:
- Accepted Manuscript
- Journal Name:
- Nanoscale
- Additional Journal Information:
- Journal Volume: 14; Journal Issue: 10; Journal ID: ISSN 2040-3364
- Publisher:
- Royal Society of Chemistry
- Country of Publication:
- United States
- Language:
- English
- Subject:
- 37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CHEMISTRY; Chemistry; Science & Technology - Other Topics; Materials Science; Physics
Citation Formats
Ren, Simiao, Mahendra, Ashwin, Khatib, Omar, Deng, Yang, Padilla, Willie J., and Malof, Jordan M. Inverse deep learning methods and benchmarks for artificial electromagnetic material design. United States: N. p., 2022.
Web. doi:10.1039/d1nr08346e.
Ren, Simiao, Mahendra, Ashwin, Khatib, Omar, Deng, Yang, Padilla, Willie J., & Malof, Jordan M. Inverse deep learning methods and benchmarks for artificial electromagnetic material design. United States. https://doi.org/10.1039/d1nr08346e
Ren, Simiao, Mahendra, Ashwin, Khatib, Omar, Deng, Yang, Padilla, Willie J., and Malof, Jordan M. Mon .
"Inverse deep learning methods and benchmarks for artificial electromagnetic material design". United States. https://doi.org/10.1039/d1nr08346e. https://www.osti.gov/servlets/purl/1978815.
@article{osti_1978815,
title = {Inverse deep learning methods and benchmarks for artificial electromagnetic material design},
author = {Ren, Simiao and Mahendra, Ashwin and Khatib, Omar and Deng, Yang and Padilla, Willie J. and Malof, Jordan M.},
abstractNote = {In this work we investigate the use of deep inverse models (DIMs) for designing artificial electromagnetic materials (AEMs) – such as metamaterials, photonic crystals, and plasmonics – to achieve some desired scattering properties (e.g., transmission or reflection spectrum). DIMs are deep neural networks (i.e., deep learning models) that are specially-designed to solve ill-posed inverse problems. There has recently been tremendous growth in the use of DIMs for solving AEM design problems however there has been little comparison of these approaches to examine their absolute and relative performance capabilities. In this work we compare eight state-of-the-art DIMs on three unique AEM design problems, including two models that are novel to the AEM community. Our results indicate that DIMs can rapidly produce accurate designs to achieve a custom desired scattering on all three problems. Although no single model always performs best, the Neural-Adjoint approach achieves the best overall performance across all problem settings. As a final contribution we show that not all AEM design problems are ill-posed, and in such cases a conventional deep neural network can perform better than DIMs. We recommend that a deep neural network is always employed as a simple baseline approach when addressing AEM design problems. Furthermore, we publish python code for our AEM simulators and our DIMs to enable easy replication of our results, and benchmarking of new DIMs by the AEM community.},
doi = {10.1039/d1nr08346e},
journal = {Nanoscale},
number = 10,
volume = 14,
place = {United States},
year = {Mon Feb 28 00:00:00 EST 2022},
month = {Mon Feb 28 00:00:00 EST 2022}
}
Free Publicly Available Full Text
Publisher's Version of Record
Other availability
Search WorldCat to find libraries that may hold this journal
Save to My Library
You must Sign In or Create an Account in order to save documents to your library.
Works referenced in this record:
Designing nanophotonic structures using conditional deep convolutional generative adversarial networks
journal, June 2019
- So, Sunae; Rho, Junsuk
- Nanophotonics, Vol. 8, Issue 7
Machine learning and evolutionary algorithm studies of graphene metamaterials for optimized plasmon-induced transparency
journal, January 2020
- Zhang, Tian; Liu, Qi; Dan, Yihang
- Optics Express, Vol. 28, Issue 13
Mapping the design space of photonic topological states via deep learning
journal, January 2020
- Singh, Robin; Agarwal, Anu; W. Anthony, Brian
- Optics Express, Vol. 28, Issue 19
Deep learning enabled inverse design in nanophotonics
journal, February 2020
- So, Sunae; Badloe, Trevon; Noh, Jaebum
- Nanophotonics, Vol. 9, Issue 5
A Bidirectional Deep Neural Network for Accurate Silicon Color Design
journal, November 2019
- Gao, Li; Li, Xiaozhong; Liu, Dianjing
- Advanced Materials, Vol. 31, Issue 51
Machine learning inverse problem for topological photonics
journal, September 2018
- Pilozzi, Laura; Farrelly, Francis A.; Marcucci, Giulia
- Communications Physics, Vol. 1, Issue 1
Deep Learning: A Rapid and Efficient Route to Automatic Metasurface Design
journal, April 2019
- Qiu, Tianshuo; Shi, Xin; Wang, Jiafu
- Advanced Science, Vol. 6, Issue 12
Design by neural network of concentric multilayered cylindrical metamaterials
journal, March 2020
- Akashi, Naoto; Toma, Mana; Kajikawa, Kotaro
- Applied Physics Express, Vol. 13, Issue 4
Study of optimization of an LCD light guide plate with neural network and genetic algorithm
journal, January 2009
- Li, Chen-Jung; Fang, Yi-Chin; Cheng, Ming-Chia
- Optics Express, Vol. 17, Issue 12
Deep learning in nano-photonics: inverse design and beyond
journal, January 2021
- Wiecha, Peter R.; Arbouet, Arnaud; Girard, Christian
- Photonics Research, Vol. 9, Issue 5
Self-Learning Perfect Optical Chirality via a Deep Neural Network
journal, November 2019
- Li, Yu; Xu, Youjun; Jiang, Meiling
- Physical Review Letters, Vol. 123, Issue 21
Deep Convolutional Mixture Density Network for Inverse Design of Layered Photonic Structures
journal, September 2020
- Unni, Rohit; Yao, Kan; Zheng, Yuebing
- ACS Photonics, Vol. 7, Issue 10
Plasmonic nanoparticle simulations and inverse design using machine learning
journal, January 2019
- He, Jing; He, Chang; Zheng, Chao
- Nanoscale, Vol. 11, Issue 37
Metamaterial-inspired silicon nanophotonics
journal, April 2017
- Staude, Isabelle; Schilling, Jörg
- Nature Photonics, Vol. 11, Issue 5
Nanophotonic particle simulation and inverse design using artificial neural networks
journal, June 2018
- Peurifoy, John; Shen, Yichen; Jing, Li
- Science Advances, Vol. 4, Issue 6
Machine learning–assisted global optimization of photonic devices
journal, October 2020
- Kudyshev, Zhaxylyk A.; Kildishev, Alexander V.; Shalaev, Vladimir M.
- Nanophotonics, Vol. 10, Issue 1
Deep learning for accelerated all-dielectric metasurface design
journal, January 2019
- Nadell, Christian C.; Huang, Bohao; Malof, Jordan M.
- Optics Express, Vol. 27, Issue 20
Optimization of photonic crystal nanocavities based on deep learning
journal, January 2018
- Asano, Takashi; Noda, Susumu
- Optics Express, Vol. 26, Issue 25
Prediction Network of Metamaterial with Split Ring Resonator Based on Deep Learning
journal, April 2020
- Hou, Zheyu; Tang, Tingting; Shen, Jian
- Nanoscale Research Letters, Vol. 15, Issue 1
Fast design of plasmonic metasurfaces enabled by deep learning
journal, October 2020
- Mall, Abhishek; Patil, Abhijeet; Tamboli, Dipesh
- Journal of Physics D: Applied Physics, Vol. 53, Issue 49
Benchmarking deep learning-based models on nanophotonic inverse design problems
journal, January 2022
- Ma, Taigao; Tobah, Mustafa; Wang, Haozhu
- Opto-Electronic Science, Vol. 1, Issue 1
Computational design of efficient grating couplers using artificial intelligence
journal, March 2020
- Miyatake, Yuto; Sekine, Naoki; Toprasertpong, Kasidit
- Japanese Journal of Applied Physics, Vol. 59, Issue SG
Deep Learning the Electromagnetic Properties of Metamaterials—A Comprehensive Review
journal, May 2021
- Khatib, Omar; Ren, Simiao; Malof, Jordan
- Advanced Functional Materials, Vol. 31, Issue 31
A Hybrid Strategy for the Discovery and Design of Photonic Structures
journal, March 2020
- Liu, Zhaocheng; Raju, Lakshmi; Zhu, Dayu
- IEEE Journal on Emerging and Selected Topics in Circuits and Systems, Vol. 10, Issue 1
Deep Learning for the Inverse Design of Mid-Infrared Graphene Plasmons
journal, February 2020
- Phan, Anh D.; Nguyen, Cuong V.; Linh, Pham T.
- Crystals, Vol. 10, Issue 2
Metasurface inverse design using machine learning approaches
journal, May 2020
- Shi, Xin; Qiu, Tianshuo; Wang, Jiafu
- Journal of Physics D: Applied Physics, Vol. 53, Issue 27
UMAP: Uniform Manifold Approximation and Projection
journal, September 2018
- McInnes, Leland; Healy, John; Saul, Nathaniel
- Journal of Open Source Software, Vol. 3, Issue 29
Probabilistic Representation and Inverse Design of Metamaterials Based on a Deep Generative Model with Semi‐Supervised Learning Strategy
journal, July 2019
- Ma, Wei; Cheng, Feng; Xu, Yihao
- Advanced Materials, Vol. 31, Issue 35
Genetically Engineered Plasmonic Nanoarrays
journal, March 2012
- Forestiere, Carlo; Pasquale, Alyssa J.; Capretti, Antonio
- Nano Letters, Vol. 12, Issue 4
Machine-learning-assisted metasurface design for high-efficiency thermal emitter optimization
journal, June 2020
- Kudyshev, Zhaxylyk A.; Kildishev, Alexander V.; Shalaev, Vladimir M.
- Applied Physics Reviews, Vol. 7, Issue 2
Genetic algorithms in engineering electromagnetics
journal, January 1997
- Johnson, J. M.; Rahmat-Samii, V.
- IEEE Antennas and Propagation Magazine, Vol. 39, Issue 4
Smart inverse design of graphene-based photonic metamaterials by an adaptive artificial neural network
journal, January 2019
- Chen, Yingshi; Zhu, Jinfeng; Xie, Yinong
- Nanoscale, Vol. 11, Issue 19
Training Deep Neural Networks for the Inverse Design of Nanophotonic Structures
journal, January 2018
- Liu, Dianjing; Tan, Yixuan; Khoram, Erfan
- ACS Photonics, Vol. 5, Issue 4
Neural-adjoint method for the inverse design of all-dielectric metasurfaces
journal, January 2021
- Deng, Yang; Ren, Simiao; Fan, Kebin
- Optics Express, Vol. 29, Issue 5
Efficient spectrum prediction and inverse design for plasmonic waveguide systems based on artificial neural networks
journal, January 2019
- Zhang, Tian; Wang, Jia; Liu, Qi
- Photonics Research, Vol. 7, Issue 3
Deep learning for the design of photonic structures
journal, October 2020
- Ma, Wei; Liu, Zhaocheng; Kudyshev, Zhaxylyk A.
- Nature Photonics, Vol. 15, Issue 2
Deep-Learning-Enabled On-Demand Design of Chiral Metamaterials
journal, May 2018
- Ma, Wei; Cheng, Feng; Liu, Yongmin
- ACS Nano, Vol. 12, Issue 6
Generative Model for the Inverse Design of Metasurfaces
journal, September 2018
- Liu, Zhaocheng; Zhu, Dayu; Rodrigues, Sean P.
- Nano Letters, Vol. 18, Issue 10
Inverse design of photonic topological state via machine learning
journal, May 2019
- Long, Yang; Ren, Jie; Li, Yunhui
- Applied Physics Letters, Vol. 114, Issue 18
Optimization of FSS with Sierpinski island fractal elements using population-based search algorithms and MLP neural network
journal, February 2014
- da Silva, Marcelo R.; Nóbrega, Clarissa de L.; Silva, Paulo H. da F.
- Microwave and Optical Technology Letters, Vol. 56, Issue 4
Deep Neural Network Inverse Design of Integrated Photonic Power Splitters
journal, February 2019
- Tahersima, Mohammad H.; Kojima, Keisuke; Koike-Akino, Toshiaki
- Scientific Reports, Vol. 9, Issue 1
Simultaneous Inverse Design of Materials and Structures via Deep Learning: Demonstration of Dipole Resonance Engineering Using Core–Shell Nanoparticles
journal, June 2019
- So, Sunae; Mun, Jungho; Rho, Junsuk
- ACS Applied Materials & Interfaces, Vol. 11, Issue 27
Enhanced light–matter interactions in dielectric nanostructures via machine-learning approach
journal, March 2020
- Xu, Lei; Rahmani, Mohsen; Ma, Yixuan
- Advanced Photonics, Vol. 2, Issue 02
Deep learning approach based on dimensionality reduction for designing electromagnetic nanostructures
journal, February 2020
- Kiarashinejad, Yashar; Abdollahramezani, Sajjad; Adibi, Ali
- npj Computational Materials, Vol. 6, Issue 1
Multitask deep-learning-based design of chiral plasmonic metamaterials
journal, January 2020
- Ashalley, Eric; Acheampong, Kingsley; Besteiro, Lucas V.
- Photonics Research, Vol. 8, Issue 7
Plasmonic nanostructure design and characterization via Deep Learning
journal, September 2018
- Malkiel, Itzik; Mrejen, Michael; Nagler, Achiya
- Light: Science & Applications, Vol. 7, Issue 1