Advances in artificial intelligence for artificial metamaterials
Source abstract
Abstract
The 2024 Nobel Prizes in Physics and Chemistry were awarded for foundational discoveries and inventions enabling machine learning through artificial neural networks. Artificial intelligence (AI) and artificial metamaterials are two cutting-edge technologies that have shown significant advancements and applications in various fields. AI, with its roots tracing back to Alan Turing’s seminal work, has undergone remarkable evolution over decades, with key advancements including the Turing Test, expert systems, deep learning, and the emergence of multimodal AI models. Electromagnetic wave control, critical for scientific research and industrial applications, has been significantly broadened by artificial metamaterials. This review explores the synergistic integration of AI and artificial metamaterials, emphasizing how AI accelerates the design and functionality of artificial materials, while novel physical neural networks constructed from artificial metamaterials significantly enhance AI’s computational speed and its ability to solve complex physical problems. This paper provides a detailed discussion of AI-based forward prediction and inverse design principles and applications in metamaterial design. It also examines the potential of big-data-driven AI methods in addressing challenges in metamaterial design. In addition, this review delves into the role of artificial metamaterials in advancing AI, focusing on the progress of electromagnetic physical neural networks in optics, terahertz, and microwaves. Emphasizing the transformative impact of the intersection between AI and artificial metamaterials, this review underscores significant improvements in efficiency, accuracy, and applicability. The collaborative development of AI and artificial metamaterials accelerates the metamaterial design process and opens new possibilities for innovations in photonics, communications, radars, and sensing.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Neural Networks and Reservoir Computing · Advanced Memory and Neural Computing · Metamaterials and Metasurfaces Applications
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Beijing Key Laboratory of Millimeter Wave and Terahertz Technology, School of Integrated Circuits and Electronics, Beijing Institute of Technology 1 , Beijing 100081,
Department of Electronic Engineering, Shanghai Jiao Tong University 2 , Shanghai 200240,
References (100+)
Show source-provided references
1. Multidimensional phase singularities in nanophotonics10.1126/science.abj0039 ↗
2. Advances in communications using optical vortices10.1364/prj.4.000b14 ↗
3. High-purity orbital angular momentum states from a visible metasurface laser10.1038/s41566-020-0623-z ↗
4. Recent advances in metasurface design and quantum optics applications with machine learning, physics-informed neural networks, and topology optimization methods10.1038/s41377-023-01218-y ↗
5. Miniaturized leaky-wave antenna with backward-to-forward beam scanning and suppressed open stop-band based on substrate-integrated plasmonic waveguide10.1364/oe.525260 ↗
6. Study of symmetries of chiral metasurfaces for azimuth-rotation-independent cross polarization conversion10.1364/oe.452773 ↗
7. Triple-band terahertz chiral metasurface for spin-selective absorption and reflection phase manipulation10.3390/electronics11244195 ↗
8. Wideband high-efficiency scattering reduction in a graphene based optically transparent and flexible metasurface10.1016/j.carbon.2024.119150 ↗
9. A mid-IR tunable graphene metasurface for ultrasensitive molecular fingerprint retrieval and refractive index sensing10.1039/d3tc03554a ↗
10. Spin valve as THz emitter providing amplitude modulation10.1063/5.0221982 ↗
11. Mechanical intelligence via fully reconfigurable elastic neuromorphic metasurfaces10.1063/5.0201761 ↗
12. Metasurface on integrated photonic platform: From mode converters to machine learning10.1515/nanoph-2022-0294 ↗
13. Electromagnetic metamaterials: From classical to quantum10.23919/emsci.2022.0005 ↗
14. Broadband extremely close-spaced 5G mimo antenna with mutual coupling reduction using metamaterial-inspired superstrate10.1364/oe.27.003472 ↗
15. Experimental demonstration of a magnetically tunable ferrite based metamaterial absorber10.1364/oe.22.016408 ↗
16. Active microwave negative-index metamaterial transmission line with gain10.1103/physrevlett.107.205503 ↗
17. Classical and generalized geometric phase in electromagnetic metasurfaces10.3788/pi.2022.r03 ↗
18. A compact quad-square negative-index metamaterial: Design, simulation, and experimental validation for microwave applications10.1063/5.0204501 ↗
19. Topology optimization of the azimuth-rotation-independent polarization conversion metasurface for bandwidth enhancement10.1364/oe.476676 ↗
20. Rapid customized design of a conformal optical transparent metamaterial absorber based on the circuit analog optimization method10.1364/oe.452694 ↗
21. Design, material, function, and fabrication of metamaterials10.1063/5.0144454 ↗
22. Experimental realization of a transmissive microwave metasurface for dual vector vortex beams generation10.1364/oe.522716 ↗
23. Tunable broadband terahertz graphene metasurface for complex-amplitude vortex beam generator and hologram10.1016/j.optlastec.2024.110874 ↗
24. Realization of terahertz frequency selecting based on topological edge states with kagome photonic crystals10.1016/j.rinp.2024.107399 ↗
25. Graphene-based ultralow-profile microwave Fresnel lens10.1016/j.carbon.2023.118599 ↗
26. High-efficiency dual-band metasurface with independent multifold geometric phases10.1002/adom.202300347 ↗
27. Development of a compact subnanosecond pulse transmitter with a single low-power supply10.1109/lmwt.2023.3279894 ↗
28. Multi-wavelength metasurface empowered cryptography for heightened security and improved fidelity10.1002/lpor.202301014 ↗
29. Electromagnetic reprogrammable coding-metasurface holograms10.1038/s41467-017-00164-9 ↗
30. Highly compact time-domain borehole radar with time-varying gain equivalent sampling10.1109/tmtt.2024.3386502 ↗
31. Light propagation with phase discontinuities: Generalized laws of reflection and refraction10.1126/science.1210713 ↗
32. Data-driven design for metamaterials and multiscale systems: A review10.1002/adma.202305254 ↗
33. Ai-based metamaterial design10.1021/acsami.4c04486 ↗
34. Unleashing the potential: Ai empowered advanced metasurface research10.1515/nanoph-2023-0759 ↗
35. Liquid crystal-integrated metasurfaces for an active photonic platform10.29026/oea.2024.230216 ↗
36. 3-D/4-D-printed reconfigurable metasurfaces for controlling electromagnetic waves10.1109/jproc.2024.3391232 ↗
37. Dual-band complex-amplitude metasurface empowered high security cryptography with ultra-massive encodable patterns10.1515/nanoph-2024-0314 ↗
38. Dual-frequency arbitrary-plane multiplexed holography at microwave frequency bands10.1002/adom.202202501 ↗
39. Mechanically reconfigurable folded reflectarray antenna for variable near-field focusing10.1109/tap.2023.3323092 ↗
40. Diffractive optical computing in free space10.1038/s41467-024-45982-w ↗
41. Deep learning in mechanical metamaterials: From prediction and generation to inverse design10.1002/adma.202302530 ↗
42. Deep learning meets nanophotonics: A generalized accurate predictor for near fields and far fields of arbitrary 3D nanostructures10.1021/acs.nanolett.9b03971 ↗
43. High-efficiency transmissive tunable metasurfaces for binary cascaded diffractive layers10.1109/tap.2024.3385112 ↗
44. Realization of high-performance optical metasurfaces over a large area: A review from a design perspective10.1038/s44310-024-00029-2 ↗
45. Out-of-distribution domain exploration by a multifidelity deep learning model to estimate electromagnetic responses of metasurfaces10.1109/tap.2024.3426290 ↗
46. Machine learning aided design and optimization of thermal metamaterials10.1021/acs.chemrev.3c00708 ↗
47. Flexible metasurfaces for multifunctional interfaces10.1021/acsnano.3c09310 ↗
48. Inverse design of metamaterial bound states in the continuum device via coupled mode theory10.1109/tmtt.2023.3305155 ↗
49. Preprocessing-based fast design of multiple EM structures with one deep neural network10.1109/tap.2024.3381376 ↗
50. Space-time-coding digital metasurface element design based on state recognition and mapping methods with CNN-LSTM-DNN10.1109/tap.2024.3349778 ↗
51. Assembling reconfigurable intelligent metasurfaces with a synthetic neural network10.1109/tap.2024.3395909 ↗
52. Transmissive Pancharatnam-Berry metasurfaces with stable amplitude and precise phase modulations using dartboard discretization configuration10.1364/oe.501702 ↗
53. Rapid design of hybrid mechanism metasurface with random coding for terahertz dual-band RCS reduction10.1364/oe.496423 ↗
54. Automatic impedance matching with a neural network10.1109/75.242258 ↗
55. Reverse modeling of microwave circuits with bidirectional neural network models10.1109/22.721152 ↗
56. Design and optimization of CPW circuits using EM-ANN models for CPW components10.1109/22.643868 ↗
57. Space mapping: The state of the art10.1109/tmtt.2003.820904 ↗
58. Blending PSO and ANN for optimal design of FSS filters with Koch island patch elements10.1109/tmag.2010.2044147 ↗
59. Machine-learning designs of anisotropic digital coding metasurfaces10.1002/adts.201800132 ↗
60. A hybrid algorithm for electromagnetic optimization utilizing neural networks
61. Artificial neural networks for rf and microwave design-from theory to practice10.1109/tmtt.2003.809179 ↗
62. Prediction of dispersion relation and PBGS in 2-D PCS by using artificial neural networks10.1109/lpt.2012.2215846 ↗
63. Novel neural network based optimization approach for photonic devices10.1007/s11082-013-9869-8 ↗
64. Artificial neural network modeling of plasmonic transmission lines10.1364/ao.55.002780 ↗
65. Machine learning approach for computing optical properties of a photonic crystal fiber10.1364/oe.27.036414 ↗
66. Computing optical properties of photonic crystals by using multilayer perceptron and extreme learning machine10.1109/jlt.2018.2856364 ↗
67. Optimization of photonic crystal nanocavities based on deep learning10.1364/oe.26.032704 ↗
68. Towards an integrated evolutionary strategy and artificial neural network computational tool for designing photonic coupler devices10.1016/j.asoc.2017.12.043 ↗
69. Efficient spectrum prediction and inverse design for plasmonic waveguide systems based on artificial neural networks10.1364/prj.7.000368 ↗
70. Deep neural network inverse design of integrated photonic power splitters10.1038/s41598-018-37952-2 ↗
71. Designing integrated photonic devices using artificial neural networks10.1364/oe.27.029620 ↗
72. An open-source artificial neural network model for polarization-insensitive silicon-on-insulator subwavelength grating couplers10.1109/JSTQE.2018.2885486 ↗
73. Nanophotonic particle simulation and inverse design using artificial neural networks10.1126/sciadv.aar4206 ↗
74. Deep-learning-enabled on-demand design of chiral metamaterials10.1021/acsnano.8b03569 ↗
75. Self-learning perfect optical chirality via a deep neural network10.1103/physrevlett.123.213902 ↗
76. Plasmonic nanostructure design and characterization via deep learning10.1038/s41377-018-0060-7 ↗
77. Finding the optical properties of plasmonic structures by image processing using a combination of convolutional neural networks and recurrent neural networks10.1038/s41378-019-0069-y ↗
78. Plasmonic colours predicted by deep learning10.1038/s41598-019-44522-7 ↗
79. Training artificial neural network for optimization of nanostructured VO2-based smart window performance10.1364/oe.27.0a1030 ↗
80. Teaching optics to a machine learning network10.1364/ol.390600 ↗
81. Neural network based design of metagratings10.1063/1.5033327 ↗
82. A deep learning approach for objective-driven all-dielectric metasurface design10.1021/acsphotonics.9b00966 ↗
83. Smart inverse design of graphene-based photonic metamaterials by an adaptive artificial neural network10.1039/c9nr01315f ↗
84. A bidirectional deep neural network for accurate silicon color design10.1002/adma.201905467 ↗
85. Training deep neural networks for the inverse design of nanophotonic structures10.1021/acsphotonics.7b01377 ↗
86. Machine learning inverse problem for topological photonics10.1038/s42005-018-0058-8 ↗
87. Mapping the design space of photonic topological states via deep learning10.1364/oe.398926 ↗
88. Application of nanogenerators in acoustics based on artificial intelligence and machine learning10.1063/5.0195399 ↗
89. Material discovery and modeling acceleration via machine learning10.1063/5.0230677 ↗
90. Fourier-space diffractive deep neural network10.1103/physrevlett.123.023901 ↗
91. Unidirectional imaging using deep learning–designed materials10.1126/sciadv.adg1505 ↗
92. Nanophotonic media for artificial neural inference10.1364/prj.7.000823 ↗
93. On-chip photonic diffractive optical neural network based on a spatial domain electromagnetic propagation model10.1364/oe.435183 ↗
94. Experimental realization of arbitrary activation functions for optical neural networks10.1364/oe.391473 ↗
95. A single layer neural network implemented by a 4 × 4 MZI-based optical processor10.1109/jphot.2019.2952562 ↗
96. Computing machinery and intelligence (1950)10.7551/mitpress/12274.003.0016 ↗
97. P. J. Werbos , “Beyond regression: New tools for prediction and analysis in the behavioral sciences,” Ph.D. thesis (Harvard University, 1974).
98. Reducing the dimensionality of data with neural networks10.1126/science.1127647 ↗
99. Generative adversarial nets
100. Mastering the game of go with deep neural networks and tree search10.1038/nature16961 ↗