Abstract

This paper proposes a deep-learning-assisted design method for 2-bit coding metasurfaces. This method uses a skip connection module and the idea of an attention mechanism in squeeze-and-excitation networks based on a fully connected network and a convolutional neural network. The accuracy limit of the basic model is further improved. The convergence ability of the model increased nearly 10 times, and the mean-square error loss function converges to 0.000168. The forward prediction accuracy of the deep-learning-assisted model is 98%, and the accuracy of inverse design results is 97%. This approach offers the advantages of an automatic design process, high efficiency, and low computational cost. It can serve users who lack metasurface design experience.

© 2023 Optica Publishing Group

Research on a demand design method of a cross polarization converter metasurface based on a depth generation model

Junwei Li, Qinhua A, Qiusong Lan, Jintao Yang, Lijun Yun, Yuelong Xia, and Chengfu Yang
Opt. Mater. Express 13(9) 2497-2510 (2023)

Deep-learning-assisted designing chiral terahertz metamaterials with asymmetric transmission properties

Feng Gao, Zhen Zhang, Yafei Xu, Liuyang Zhang, Ruqiang Yan, and Xuefeng Chen
J. Opt. Soc. Am. B 39(6) 1511-1519 (2022)

Dynamic multifunctional metasurfaces: an inverse design deep learning approach

Zhi-Dan Lei, Yi-Duo Xu, Cheng Lei, Yan Zhao, and Du Wang
Photon. Res. 12(1) 123-133 (2024)

Deep learning for the design of 3D chiral plasmonic metasurfaces

Xianglai Liao, Lili Gui, Zhenming Yu, Tian Zhang, and Kun Xu
Opt. Mater. Express 12(2) 758-771 (2022)

Deep-learning based broadband reflection reduction metasurface

Haiyan Xie, Xiuli Yue, Kaihuai Wen, Difei Liang, Tiancheng Han, and Longjiang Deng
Opt. Express 31(9) 14593-14603 (2023)