A Bidirectional Deep Neural Network for Accurate Silicon Color Design

TLDR

Silicon nanostructure color achieves unprecedented high printing resolution and larger color gamut than sRGB, but designing specific colors is computationally costly due to the sensitivity of localized magnetic and electric dipole resonances to geometry. The study trains a deep neural network to predict the color of random silicon nanostructures and to solve the inverse design nonuniqueness problem. The network accurately models forward color generation and outputs device geometries for at least one million distinct colors. Deep learning dramatically reduces computation cost and increases design efficiency, enabling high‑accuracy silicon color manufacturing.

Abstract

Silicon nanostructure color has achieved unprecedented high printing resolution and larger color gamut than sRGB. The exact color is determined by localized magnetic and electric dipole resonance of nanostructures, which are sensitive to their geometric changes. Usually, the design of specific colors and iterative optimization of geometric parameters are computationally costly, and obtaining millions of different structural colors is challenging. Here, a deep neural network is trained, which can accurately predict the color generated by random silicon nanostructures in the forward modeling process and solve the nonuniqueness problem in the inverse design process that can accurately output the device geometries for at least one million different colors. The key results suggest deep learning is a powerful tool to minimize the computation cost and maximize the design efficiency for nanophotonics, which can guide silicon color manufacturing with high accuracy.