Abstract:A global design strategy based on field‐driven polygon evolution is proposed to achieve the inverse design of large‐scale coupled meta‐atoms. Through two global simulations, it can effectively reshape any given target optical field into an optimal structural distribution of devices without knowing mapping relationship. During the past few years, metasurface polarization optics has experienced remarkable advances, resulting in revolutionary applications in imaging, sensing, computing, etc. The realization of complex optical operations requires the consideration of both the individual meta‐atoms as well as their intricate couplings. However, conventional design methods face challenges as design degrees of freedom and functionality complexity. Additionally, previous studies are restricted to the local design of single meta‐atoms based on explicit mapping relationships while ignoring interactions, resulting in an inability to meet the on‐demand requirements of complex light‐field operations. Here, a global design strategy based on field‐driven polygon evolution to achieve the inverse design of large‐scale coupled meta‐atoms is proposed. Through two global simulations, it can effectively reshape any given target optical field into an optimal structural distribution of devices without knowing mapping relationship. Near‐perfect spin‐decoupled beam‐splitting and high‐performance focusing, as well as the generation of arbitrary vector optical fields on the Poincaré sphere with a maximal diffraction efficiency closely approaching 100%, are experimentally demonstrated. This strategy opens up a new avenue for a rapid inverse design of large‐scale, high‐performance multifunctional meta‐devices, which can hold significant implications for both classical and quantum information processing domains.

Inverse design and experimental realization of plasma metamaterials

Inverse Design of Plasma Metamaterial Devices for Optical Computing

Neural Network-Based Surrogate Model for Inverse Design of Metasurfaces

Experimental Characterization of the Dispersive Behavior in a Uniaxial Metamaterial Around Plasma Frequency.

Inverse Design and Experimental Verification of a Bianisotropic Metasurface Using Optimization and Machine Learning

Physics-Informed Inverse Design of Programmable Metasurfaces

Designing Metasurfaces to Manipulate Antenna Radiation

Inverse Design of Two-Function Transmission-type Reconfigurable Polarization Control Metasurfaces Based on Deep Learning

Inverse Design of an All-Dielectric Nonlinear Polaritonic Metasurface

Inverse Design of Discrete Mechanical Metamaterials

Physics-informed Inverse Design of Multi-bit Programmable Metasurfaces

Field‐Driven Inverse Design of High‐Performance Polarization‐Multiplexed Meta‐devices

Deep learning to accelerate Maxwell's equations for inverse design of dielectric metasurfaces

Machine-Learning-Based Characterization and Inverse Design of Metamaterials

A Combined Machine-Learning / Optimization-Based Approach for Inverse Design of Nonuniform Bianisotropic Metasurfaces

Inverse design of electromagnetic metamaterials: From iterative to deep learning-based methods

Inverse design of a 1D dielectric metasurface by topology optimization: fluctuations-trend analysis assisted by a diamond-square algorithm

Inverse Design of Metamaterials via Deep Learning for Electromagnetically Induced Transparency

Experimental Realization of an Achromatic Magnetic Mirror based on Metamaterials

Inverse design of transmission-type linear-to-circular polarization control metasurface based on deep learning

Inverse design of phononic meta-structured materials