We show that generative neural networks can train from images of periodic, topology-optimized metasurfaces to produce high-efficiency, topologically complex devices, served as a new design tool to facilitate metasurface design.

© 2019 The Author(s)

PDF Article


You do not have subscription access to this journal. Citation lists with outbound citation links are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
Login to access OSA Member Subscription