Inverse Design of Diffractive Metasurfaces Using Diffusion Models

TL;DR

This paper introduces a diffusion model-based method for inverse design of metasurfaces, enabling efficient and accurate creation of optical elements with complex desired scattering properties.

Contribution

The work integrates diffusion models into metasurface inverse design, providing a data-driven approach that reduces computational costs and improves design accuracy.

Findings

Achieved low-error metasurface designs in under 30 minutes.

Successfully designed spatially uniform intensity and polarization beam splitters.

Provided publicly available code and datasets for further research.

Abstract

Metasurfaces are ultra-thin optical elements composed of engineered sub-wavelength structures that enable precise control of light. Their inverse design - determining a geometry that yields a desired optical response - is challenging due to the complex, nonlinear relationship between structure and optical properties. This often requires expert tuning, is prone to local minima, and involves significant computational overhead. In this work, we address these challenges by integrating the generative capabilities of diffusion models into computational design workflows. Using an RCWA simulator, we generate training data consisting of metasurface geometries and their corresponding far-field scattering patterns. We then train a conditional diffusion model to predict meta-atom geometry and height from a target spatial power distribution at a specified wavelength, sampled from a continuous supported band. Once trained, the model can generate metasurfaces with low error, either directly using RCWA-guided posterior sampling or by serving as an initializer for traditional optimization methods. We demonstrate our approach on the design of a spatially uniform intensity splitter and a polarization beam splitter, both produced with low error in under 30 minutes. To support further research in data-driven metasurface design, we publicly release our code and datasets.