Abstract

In this Letter, a transfer learning method is proposed to complete design tasks on heterogeneous metasurface datasets with distinct functionalities. Through fine-tuning the inverse design network and freezing the parameters of hidden layers, we successfully transfer the metasurface inverse design knowledge from the electromagnetic-induced transparency (EIT) domain to the three target domains of EIT (different design), absorption, and phase-controlled metasurface. Remarkably, in comparison to the source domain dataset, a minimum of only 700 target domain samples is required to complete the training process. This work presents a significant solution to lower the data threshold for the inverse design process and provides the possibility of knowledge transfer between different domain metasurface datasets.

© 2024 Optica Publishing Group

Deep learning denoising diffusion probabilistic model applied to holographic data synthesis

Alejandro Velez-Zea, Cristian David Gutierrez-Cespedes, and John Fredy Barrera-Ramírez
Opt. Lett. 49(3) 514-517 (2024)

Full Stokes polarimetry based on an inverse-designed multi-foci metalens

Shuyi Wang, Tie Hu, Shichuan Wang, Yunxuan Wei, Zihan Mei, Bing Yan, Wenhong Zhou, Zhenyu Yang, JinKun Zheng, YuanLong Peng, and Ming Zhao
Opt. Lett. 49(6) 1595-1598 (2024)

Deep learning approach for inverse design of metasurfaces with a wider shape gamut

Soumyashree S. Panda, Sumit Choudhary, Siddharth Joshi, Satinder K. Sharma, and Ravi S. Hegde
Opt. Lett. 47(10) 2586-2589 (2022)

Physics-based supervised learning method for high dynamic range 3D measurement with high fidelity

Fuqian Li, Xingman Niu, Jing Zhang, Qican Zhang, and Yajun Wang
Opt. Lett. 49(3) 602-605 (2024)

Monolayer directional metasurface for all-optical image classifier doublet

Rui Xia, Lin Wu, Jin Tao, Ming Zhao, and ZhenYu Yang
Opt. Lett. 49(9) 2505-2508 (2024)