Abstract:The research of metamaterials has achieved enormous success in the manipulation of light in an artificially prescribed manner using delicately designed sub-wavelength structures, so-called meta-atoms. Even though modern numerical methods allow to accurately calculate the optical response of complex structures, the inverse design of metamaterials is still a challenging task due to the non-intuitive and non-unique relationship between physical structures and optical responses. To better unveil this implicit relationship and thus facilitate metamaterial design, we propose to represent metamaterials and model the inverse design problem in a probabilistically generative manner. By employing an encoder-decoder configuration, our deep generative model compresses the meta-atom design and optical response into a latent space, where similar designs and similar optical responses are automatically clustered together. Therefore, by sampling in the latent space, the stochastic latent variables function as codes, from which the candidate designs are generated upon given requirements in a decoding process. With the effective latent representation of metamaterials, we can elegantly model the complex structure-performance relationship in an interpretable way, and solve the one-to-many mapping issue that is intractable in a deterministic model. Moreover, to alleviate the burden of numerical calculation in data collection, we develop a semi-supervised learning strategy that allows our model to utilize unlabeled data in addition to labeled data during training, simultaneously optimizing the generative inverse design and deterministic forward prediction in an end-to-end manner. On a data-driven basis, the proposed model can serve as a comprehensive and efficient tool that accelerates the design, characterization and even new discovery in the research domain of metamaterials and photonics in general.

Inverse Design of Semiconductor Materials with Deep Generative Models

Generative deep learning for the inverse design of materials

Generative Deep Neural Networks for Inverse Materials Design Using Backpropagation and Active Learning

Generative Models for Inverse Design of Inorganic Solid Materials

Generative deep learning as a tool for inverse design of high-entropy refractory alloys

Inverse design of two-dimensional materials with invertible neural networks

A Generative Model for Inverse Design of Metamaterials

Deep Generative Models for Materials Discovery and Machine Learning-Accelerated Innovation

Global Inverse Design Across Multiple Photonic Structure Classes Using Generative Deep Learning

An Invertible, Invariant Crystal Representation for Inverse Design of Solid-State Materials Using Generative Deep Learning

Inverse Materials Design of Doping Strategies with AI, Thermodynamics, and Density Functional Theory

Generative adversarial networks (GAN) based efficient sampling of chemical composition space for inverse design of inorganic materials

Generative adversarial networks (GAN) based efficient sampling of chemical space for inverse design of inorganic materials

Active learning based generative design for the discovery of wide bandgap materials

Deep-learning-based inverse design model for intelligent discovery of organic molecules

Probabilistic representation and inverse design of metamaterials based on a deep generative model with semi-supervised learning strategy

PSP-GEN: Stochastic inversion of the Process-Structure-Property chain in materials design through deep, generative probabilistic modeling

An invertible crystallographic representation for general inverse design of inorganic crystals with targeted properties

Inverse Design for Materials Discovery from the Multidimensional Electronic Density of States

Deep Reinforcement Learning for Inverse Inorganic Materials Design

Tailoring the mechanical properties of 3D microstructures: a deep learning and genetic algorithm inverse optimization framework