A data-efficient self-supervised deep learning model for design and characterization of nanophotonic structures
- PDF / 4,864,080 Bytes
- 8 Pages / 595.276 x 793.701 pts Page_size
- 5 Downloads / 179 Views
gust 2020 Vol. 63 No. 8: 284212 https://doi.org/10.1007/s11433-020-1575-2
Special Topic: Metasurfaces of Novel Designs and Functionalities
A data-efficient self-supervised deep learning model for design and characterization of nanophotonic structures 1
Wei Ma , and Yongmin Liu 1 2
2,3*
College of Information Science and Electronic Engineering, Zhejiang University, Hangzhou 310027, China; Department of Mechanical and Industrial Engineering, Northeastern University, Massachusetts 02115, USA; 3 Department of Electrical and Computer Engineering, Northeastern University, Massachusetts 02115, USA Received March 3, 2020; accepted May 7, 2020; published online June 22, 2020
With its tremendous success in many machine learning and pattern recognition tasks, deep learning, as one type of data-driven models, has also led to many breakthroughs in other disciplines including physics, chemistry and material science. Nevertheless, the supremacy of deep learning over conventional optimization approaches heavily depends on the huge amount of data collected in advance to train the model, which is a common bottleneck of such a data-driven technique. In this work, we present a comprehensive deep learning model for the design and characterization of nanophotonic structures, where a self-supervised learning mechanism is introduced to alleviate the burden of data acquisition. Taking reflective metasurfaces as an example, we demonstrate that the self-supervised deep learning model can effectively utilize randomly generated unlabeled data during training, with the total test loss and prediction accuracy improved by about 15% compared with the fully supervised counterpart. The proposed self-supervised learning scheme provides an efficient solution for deep learning models in some physics-related tasks where labeled data are limited or expensive to collect. nanophotonics, metasurfaces, self-supervised, deep learning PACS number(s): 78.67.-n, 78.20.Bh, 07.05.Mh Citation:
1
W. Ma, and Y. Liu, A data-efficient self-supervised deep learning model for design and characterization of nanophotonic structures, Sci. China-Phys. Mech. Astron. 63, 284212 (2020), https://doi.org/10.1007/s11433-020-1575-2
Introduction
Nanophotonics, the study of light-matter interactions at the nanoscale, enables us to control the flow of light within the dimension far below the optical wavelength. It spawns a plethora of novel applications, such as miniaturized flat optics [1], perfect absorption [2], sub-diffraction-limited imaging [3] and extreme light concentration [4]. In all these applications, artificially designed structures play a crucial role in engineering the light-matter interactions [5,6]. Conventionally, such artificial structures as metamaterials/me*Corresponding author (email: [email protected])
tasurfaces, photonic crystals and plasmonic nano-structures are designed based on expert experience. Certain empirical templates are used as initial guess, from which a limited set of design parameters are adjusted to optimize the design, either by a
Data Loading...
