In Situ Customized Illusion Enabled by Global Metasurface Reconstruction

Abstract

Optical illusion has always attracted extensive attention, as it provides a superior self‐protection ability for both natural animals and human beings. A decade ago, this motivated the study and application of transformation optics, which provides a universal tool to manipulate light for invisibility cloaking and optical illusion. However, mainstream transformation‐optics‐based optical illusions are inherently hindered by the extreme requirements of metamaterial compositions in practice and unfavorably limited by the very large computational cost caused by their bulky state. To overcome these grand challenges, a novel and intelligent optical illusion supported by form‐free metasurfaces via a deep learning architecture is reported, which can not only render a similar illusion effect but also greatly reduces the parameter space in physics. Illustrative examples of conformal metasurfaces are presented, with a high‐fidelity inverse design from either the near‐ or far‐field in the simulation and experiment. Furthermore, a full set of intelligent systems is developed to benchmark the real‐world optical illusion applicability. The work brings the available illusion strategies closer to a wide range of in situ practical‐oriented applications and lays a foundation for the next generation of intelligent metamaterials.

Dynamic recognition and mirage using neuro-metamaterials

“…Our work provides deep insight into and a better understanding of the way metamaterials can facilitate real-world applications and may trigger other existing or yet-to-be conceived applications, in terms of wireless communication 22 , illusive cloak 23 , 39 , 40 , and information processing and display 24 , 25 . In some applications that involve high-throughput and on-the-fly data processing, low-energy consumption, and low-heat generation, neuro-metamaterials may find definite advantages over electronic counterparts, such as accelerating matrix calculation in artificial neural network 15 , passive imaging pre-processor 11 – 13 , and synthetic aperture radar 41 .…”

Homeostatic neuro-metasurfaces for dynamic wireless channel management

“…Another meaningful improvement would be the use of semisupervised and unsupervised learning, which would largely relax the high reliance on massive data collection for even larger-scale neuro-metasurfaces. For a more general open-loop operation system, the on-site learning working mode can be applied to regulate wireless channel, providing robustness to unexpected stimuli ( 53 – 54 ). In turn, we also anticipate that the homeostatic neuro-metasurfaces will accelerate deep learning algorithm in optics by harnessing the advantages of parallel computing and speed-of-light operation ( 15 ).…”

Heterogeneous Transfer‐Learning‐Enabled Diverse Metasurface Design

“…More prominently, the proposed method can be readily extended to other research domains of optics and materials science, giving rise to a multitude of on-demand photonic designs (e.g., 2D graphene and 3D topological semimetals) [33][34][35][36] and a broad range of applications, such as cloaking, imaging, and wireless communications. [3][4][5][10][11][12][13][14] We envision that interdisciplinary researchers, such as physicists, optics scientists, machine learning researchers, and materials specialists, can team up to contribute to a data and network bank covering various metasurfaces. The "global metasurface gene bank" will not only enhance the data utilization efficiency but also reduce the substantial computational burden and human labor.…”

“…[6,7] Thus far, metasurfaces have led to a myriad of exciting applications, such as planar lenses and invisibility cloaks, and have motivated scientists to revisit some recognized physical laws. [8][9][10][11][12][13][14][15] For both scientific explorations and practical-oriented studies, the first step involves designing metasurfaces. The efficiency and outcome of metasurface design largely determine the performance of the resultant metasurface devices.…”