OSA Travelling Lecture - Discovering the Electromagnetic Properties of All-Dielectric Metasurfaces with Machine Learning

Professor Willie Padilla
Department of Electrical and Computer Engineering, Duke University, USA

All-dielectric metasurfaces are a versatile platform to investigate a host of unconventional physical scattering responses. Effects, including Huygens’ surfaces, bound-states-in-the-continuum, coherent perfect absorbers, zero-rank absorbers, antireflective surfaces, and reflect arrays have been demonstrated. In principle these promising structures can be combined using supercell designs – like those in metal-based metasurfaces – in order to achieve even greater functionality. However, unlike metal-based metamaterials, structures which constitute all-dielectric metasurfaces exhibit significant unit cell interactions, thus making conventional optimization techniques impossible or impractical. Here we demonstrate modeling of complex all-dielectric metasurface systems with deep neural networks, using both the metasurface geometry and knowledge of the underlying physics as inputs. Our deep learning network is highly accurate, achieving an average mean square error of only 1.16 × 10−3 and is over five orders of magnitude faster than conventional electromagnetic simulation software. These techniques significantly increase the viability of more complex all-dielectric metasurface designs and provide opportunities for the future of tailored light matter interactions.

Willie Padilla is a Full Professor in the Department of ECE at Duke University with MS and PhD degrees in Physics from the University of California San Diego. He was a Director’s Postdoctoral Fellow at Los Alamos National Laboratory. In 2007 he was awarded a Young Investigator Award from the Office of Naval Research, and Presidential Early Career Award for Scientists and Engineers in 2011. In 2012 he was elected a Fellow of the Optical Society of America, and a Kavli Frontiers of Science Fellow in 2013. Dr. Padilla was elevated to Senior Member of the SPIE in 2018, and is a Fellow of the American Physical Society. Professor Padilla is a Web of Science Highly Cited Researcher in the field of Physics since 2018, has more than 200 peer-reviewed journal article, two book chapters, and seven issued patents. He heads a group working in the area of artificially structured systems including metamaterials with a focus on machine learning, computational imaging, spectroscopy and energy.

Sponsored by The Optical Society (OSA)

Date & time

Wed 11 Dec 2019, 11am–12pm



Seminar Room (414)


Members of RSPE welcome