arXiv subjects

Sunae So

Publications and source records attributed to Sunae So.

5 recordsLinked to original sources

arXiv2026

Machine learning classification of dynamical phases in a driven nonlinear Su-Schrieffer-Heeger lattice

Driven nonlinear photonic lattices can settle into steady, symmetry-broken, periodic, or chaotic states, and mapping where each occurs in parameter space is key to device design. Brute-force mapping is slow and relies on hand-built criteria, while machine-learning classifiers are usually tested on different systems, so it is unclear which suits a new lattice. We therefore apply four classifiers to the same coherently driven Su-Schrieffer-Heeger (SSH) chain of Kerr resonators: a hand-built heuristic classifier based on Lyapunov exponents and spectral diagnostics, a supervised convolutional neural network (CNN) trained on the heuristic labels, an unsupervised library-building scheme based on augmented dynamic mode decomposition (aDMD), and an unsupervised Gaussian mixture model (GMM) on four dynamical invariants that include an aDMD reconstruction error. Using a converged heuristic diagram as the reference, the CNN reaches 98.0% held-out accuracy, while the two label-free methods reach adjusted Rand indices (ARIs) of 0.916 and 0.947, respectively, without seeing any labels. We show that the aDMD library cleanly resolves the stable and bistable phases and separates symmetric from symmetry-broken periodic oscillations, which the heuristic misses, though it cannot distinguish periodic from chaotic switching. We also find that a poorly converged reference understates the label-free methods' accuracy.

physics.optics↗

arXiv2022

Multicolor and 3D holography realized by inverse design of single-celled metasurfaces

Metasurface-generated holography has emerged as a promising route for fully reproducing vivid scenes by manipulating the optical properties of light using ultra-compact devices. However, achieving multiple holographic images using a single metasurface is still difficult due to the capacity limit of a single meta-atom. In this work, we present an inverse design method based on gradient-descent optimization to encode multiple pieces of holographic information into a single metasurface. The proposed method allows the inverse design of single-celled metasurfaces without the need for complex meta-atom design strategies, facilitating high-throughput fabrication using broadband low loss materials. By exploiting the proposed design method, both multiplane RGB color holograms and 3D holograms are designed and experimentally demonstrated. Up to eighteen distinct metasurface-generated holographic images are demonstrated, achieving the state-of-the-art data capacity of a single phase-only metasurface. To the best of our knowledge, we present the first experimental demonstration of metasurface-generated 3D holograms that have completely independent and distinct images in each plane. By demonstrating the high-density holographic information of a single metasurface, the current research findings provide a viable route for practical metasurfaces-generated holography, ultimately stepping towards applications in optical storage, displays, and full-color imaging.

physics.optics↗

arXiv2020

Describing meta-atoms using the exact higher-order polarizability tensors

In nanophotonics, multipole framework has become an indispensable theoretical tool for analyzing subwavelength meta-atoms and their radiation properties. This work presents higher-order exact dynamic polarizability (alpha) tensors, which can fully represent anisotropic meta-atoms with higher-order multipole transitions. By using the irreducible exact Cartesian multipoles and field components as the basis, the exact alpha-tensor rigorously reflects symmetry information of particles including reciprocity. In addition, the exact alpha-tensor can be obtained from T-matrix simply using basis transformation. Finally, we show that description of meta-atoms using alpha-tensors incorporated with multiple-scattering theory vastly extends the applicability of the multipole framework in nanophotonics, allowing accurate and efficient depiction of complicated, random, multi-scale systems.

physics.optics↗

arXiv2019

Simultaneous inverse design of materials and parameters of core-shell nanoparticle via deep-learning: Demonstration of dipole resonance engineering

Recent introduction of data-driven approaches based on deep-learning technology has revolutionized the field of nanophotonics by allowing efficient inverse design methods. In this paper, simultaneous inverse design of materials and structure parameters of core-shell nanoparticle is achieved for the first time using deep-learning of a neural network. A neural network to learn correlation between extinction spectra of electric and magnetic dipoles and core-shell nanoparticle designs, which include material information and shell thicknesses, is developed and trained. We demonstrate deep-learning-assisted inverse design of core-shell nanoparticle for 1) spectral tuning electric dipole resonances, 2) finding spectrally isolated pure magnetic dipole resonances, and 3) finding spectrally overlapped electric dipole and magnetic dipole resonances. Our finding paves the way of the rapid development of nanophotonics by allowing a practical utilization of a deep-learning technology for nanophotonic inverse design.

physics.optics↗

arXiv2019

Designing nanophotonic structures using conditional-deep convolutional generative adversarial networks

Data-driven design approaches based on deep-learning have been introduced in nanophotonics to reduce time-consuming iterative simulations which have been a major challenge. Here, we report the first use of conditional deep convolutional generative adversarial networks to design nanophotonic antennae that are not constrained to a predefined shape. For given input reflection spectra, the network generates desirable designs in the form of images; this form allows suggestions of new structures that cannot be represented by structural parameters. Simulation results obtained from the generated designs agreed well with the input reflection spectrum. This method opens new avenues towards the development of nanophotonics by providing a fast and convenient approach to design complex nanophotonic structures that have desired optical properties.

physics.optics↗