Metasurface presents itself as a method to create flat optical devices that generate customizable wavefronts at the nanoscale. The traditional metasurface design process involves solving Maxwell’s equations through forward simulations and implementing trial-and-error to achieve the desired spectral response. This approach is computationally expensive and typically requires multiple iterations. In this study, we propose a reverse engineering solution that utilizes a deep learning artificial neural network (DNN). The ideal phase and transmission spectrums are inputted into the neural network, and the predicted dimensions which correspond to these spectrums are outputted by the network. The prediction process is less computationally expensive than forward simulations and is orders of magnitude faster to execute. Our neural network aims to identify the dimensions of elliptical nanopillars that will create the ideal phase response with a near unity transmission in a 20 nm wavelength interval surrounding the center wavelength of the spectral response. We have trained such a reverse DNN to predict the optimal dimensions for a birefringent metasurface composed of elliptical nanopillars.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.