Abstract
Applying a transfer-learning technique for generative deep neural networks, we demonstrate a very time-efficient inverse design framework for photonic integrated circuit devices, when there are new demands for structural/material parameters from an existing device library.
© 2022 IEEE
PDF ArticleMore Like This
Minwoo Jung, Keisuke Kojima, Toshiaki Koike-Akino, Ye Wang, Dayu Zhu, and Matthew Brand
SW5E.6 CLEO: Science and Innovations (CLEO:S&I) 2022
Otto L. Muskens
JW2D.3 Flat Optics: Components to Systems (FLATOPTICS) 2021
Toshiaki Koike-Akino, Keisuke Kojima, and Ye Wang
JW3A.44 CLEO: Applications and Technology (CLEO:A&T) 2022