By co-designing a meta-optical front end in conjunction with an image-processing back end, we demonstrate noise sensitivity and compactness substantially superior to either an optics-only or a computation-only approach, illustrated by two examples: subwavelength imaging and reconstruction of the full polarization coherence matrices of multiple light sources. Our end-to-end inverse designs couple the solution of the full Maxwell equations---exploiting all aspects of wave physics arising in subwavelength scatterers---with inverse-scattering algorithms in a single large-scale optimization involving $\gtrsim 10^4$ degrees of freedom. The resulting structures scatter light in a way that is radically different from either a conventional lens or a random microstructure, and suppress the noise sensitivity of the inverse-scattering computation by several orders of magnitude. Incorporating the full wave physics is especially crucial for detecting spectral and polarization information that is discarded by geometric optics and scalar diffraction theory.
A deep learning-based wavelength controllable forward prediction and inverse design model of nanophotonic devices is proposed. Both the target time-domain and wavelength-domain information can be utilized simultaneously, which enables multiple functions, including power splitter and wavelength demultiplexer, to be implemented efficiently and flexibly.
Data inconsistency leads to a slow training process when deep neural networks are used for the inverse design of photonic devices, an issue that arises from the fundamental property of non-uniqueness in all inverse scattering problems. Here we show that by combining forward modeling and inverse design in a tandem architecture, one can overcome this fundamental issue, allowing deep neural networks to be effectively trained by data sets that contain non-unique electromagnetic scattering instances. This paves the way for using deep neural networks to design complex photonic structures that requir…
Recent advancements in computational inverse design have begun to reshape the landscape of structures and techniques available to nanophotonics. Here, we outline a cross section of key developments at the intersection of these two fields: moving from a recap of foundational results to motivation of emerging applications in nonlinear, topological, near-field and on-chip optics.
We present a digitized adjoint method for realizing efficient inverse design of "digital" subwavelength nanophotonic devices. We design a single-mode 3-dB power divider and a dual-mode demultiplexer to demonstrate the digitized adjoint method for single-object and dual-object optimizations, respectively. The optimization comprises three stages, a first stage of continuous variation for an "analog" pattern, a second stage of forced permittivity biasing for a "quasi-digital" pattern, and a third stage for a multi-level digital pattern. Compared with conventional brute-force method, the proposed …
The generation of radiation from free electrons passing a grating, known as Smith-Purcell radiation, finds various applications including non-destructive beam diagnostics and tunable light sources, ranging from terahertz towards X-rays. So far, the gratings used for this purpose have been designed manually, based on human intuition and simple geometric shapes. Here we apply the computer-based technique of nanophotonic inverse design to build a 1400nm Smith-Purcell radiator for sub-relativistic 30 keV electrons. We demonstrate that the resulting silicon nanostructure radiates with a 3-times-hig…
Predicting physical response of an artificially structured material is of particular interest for scientific and engineering applications. Here we use deep learning to predict optical response of artificially engineered nanophotonic devices. In addition to predicting forward approximation of transmission response for any given topology, this approach allows us to inversely approximate designs for a targeted optical response. Our Deep Neural Network (DNN) could design compact (2.6x2.6 μm2) silicon-on-insulator (SOI)-based 1 X 2 power splitters with various target splitting ratios in a fraction …
The development of inverse design, where computational optimization techniques are used to design devices based on certain specifications, has led to the discovery of many compact, non-intuitive structures with superior performance. Among various methods, large-scale, gradient-based optimization techniques have been one of the most important ways to design a structure containing a vast number of degrees of freedom. These techniques are made possible by the adjoint method, in which the gradient of an objective function with respect to all design degrees of freedom can be computed using only two…
The conventional approach to nanophotonic metasurface design and optimization for a targeted electromagnetic response involves exploring large geometry and material spaces, which is computationally costly, time consuming and a highly iterative process based on trial and error. Moreover, the non-uniqueness of structural designs and high non-linearity between electromagnetic response and design makes this problem challenging. To model this non-intuitive relationship between electromagnetic response and metasurface structural design as a probability distribution in the design space, we introduce …
In recent years, the development of nanophotonic devices has presented a revolutionary means to manipulate light at nanoscale. Recently, artificial neural networks (ANNs) have displayed powerful ability in the inverse design of nanophotonic devices. However, there is limited research on the inverse design for modeling and learning the sequence characteristics of a spectrum. In this work, we propose a novel deep learning method based on an improved recurrent neural networks to extract the sequence characteristics of a spectrum and achieve inverse design and spectrum prediction. A key feature of…
Lithium niobate-on-insulator (LNOI) is an emerging photonic platform that exhibits favorable material properties (such as low optical loss, strong nonlinearities, and stability) and enables large-scale integration with stronger optical confinement, showing promise for future optical networks, quantum processors, and nonlinear optical systems. However, while photonics engineering has entered the era of automated "inverse design" via optimization in recent years, the design of LNOI integrated photonic devices still mostly relies on intuitive models and inefficient parameter sweeps, limiting the …
Dimension expansion for simulation-efficient nanophotonic neural networks
Stevens Neuroimaging and Informatics Institute, University of Southern California（南加州大学史蒂恩斯神经影像与信息学研究所）；Department of Biomedical Engineering, University of Southern California（南加州大学生物医学工程系）；Department of Electrical and Computer Engineering, University of Southern California（南加州大学电气与计算机工程系）；Yale University（耶鲁大学）
Adjoint optimization is an effective method in the inverse design of nanophotonic devices. In order to ensure the manufacturability, one would like to have control over the minimal feature sizes. Here we propose utilizing a level-set method based on b-spline surfaces in order to control the feature sizes. This approach is first used to design a wavelength demultiplexer. It is also used to implement a nanophotonic structure for artificial neural computing. In both cases, we show that the minimal feature sizes can be easily parameterized and controlled.
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-…
Gradient-Informed Bayesian and Interior Point Optimization for Efficient Inverse Design in Nanophotonics
Institute of Information Processing, Leibniz University Hannover（莱比锡洪堡大学信息处理研究所）；Flexcompute, Watertown, MA, USA（Flexcompute公司）；Institute of Artificial Intelligence, Leibniz University Hannover（莱比锡洪堡大学人工智能研究所）；L3S Research Center, Hannover, Germany（L3S研究中心）
Neural network inverse design of nanophotonic scintillators
Technion – Israel Institute of Technology（以色列理工学院）；Institute for Instrumentation in Molecular Imaging (I3M), CSIC-Universitat Politècnica de València（分子成像仪器研究所，西班牙瓦伦西亚理工大学）；E.L. Ginzton Laboratory, Stanford University（斯坦福大学）；Institute of Science and Technology Austria (ISTA)（奥地利科学技术研究院）
Max Planck Institute for the Science of Light（马克斯·普朗克光科学研究所）；Friedrich-Alexander Universität Erlangen-Nürnberg（埃尔朗根-纽伦堡大学）；Technical University of Denmark（丹麦技术大学）；John A. Paulson School of Engineering and Applied Sciences, Harvard University（哈佛大学约翰·A·保尔森工程与应用科学学院）；Massachusetts Institute of Technology（麻省理工学院）；The Chinese University of Hong Kong, Shenzhen（香港中文大学（深圳））；SPINS Photonics Inc（SPINS光子有限公司）
University of California, Los Angeles（加利福尼亚大学洛杉矶分校）；Northrop Grumman Corporation（诺斯罗普·格鲁曼公司）；California NanoSystems Institute（加州纳米系统研究所）
From higher computational efficiency to enabling the discovery of novel and complex structures, deep learning has emerged as a powerful framework for the design and optimization of nanophotonic circuits and components. However, both data-driven and exploration-based machine learning strategies have limitations in their effectiveness for nanophotonic inverse design. Supervised machine learning approaches require large quantities of training data to produce high-performance models and have difficulty generalizing beyond training data given the complexity of the design space. Unsupervised and rei…
Inverse design techniques in the context of nanophotonics have helped in discovery of compact and counter-intuitive structures/shapes. We introduce the concept of spectral domain inverse design to search through the optical trade-space (dispersive permittivity) of nanocomposite metamaterials. We develop a hybrid optimization technique that combines genetic algorithms and gradient descent methods. We utilize this technique to inverse design an ultra-thin thermophotovoltaic emitter coating material. Our work can lead to an efficient approach to search for new multi-functional optical/thermal met…
Nanophotonics has been an active research field over the past two decades, triggered by the rising interests in exploring new physics and technologies with light at the nanoscale. As the demands of performance and integration level keep increasing, the design and optimization of nanophotonic devices become computationally expensive and time-inefficient. Advanced computational methods and artificial intelligence, especially its subfield of machine learning, have led to revolutionary development in many applications, such as web searches, computer vision, and speech/image recognition. The comple…
Technion – Israel Institute of Technology（以色列理工学院）；Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU)（埃尔朗根-纽伦堡弗里德里希·亚历山大大学）；University of Cambridge（剑桥大学）
Single solid-state quantum emitters offer considerable potential for the implementation of sources of indistinguishable single-photons, which are central to many photonic quantum information systems. Nanophotonic geometry optimization with multiple performance metrics is imperative to convert a bare quantum emitter into a single-photon source that approaches the necessary levels of purity, indistinguishability, and brightness for quantum photonics. We present an inverse design methodology that simultaneously targets two important figures-of-merit for high-performance quantum light sources: the…
Karlsruhe Institute of Technology（卡尔斯鲁厄理工学院）；University of Tartu（塔尔图大学）
Neural operators have emerged as a powerful tool for solving partial differential equations in the context of scientific machine learning. Here, we implement and train a modified Fourier neural operator as a surrogate solver for electromagnetic scattering problems and compare its data efficiency to existing methods. We further demonstrate its application to the gradient-based nanophotonic inverse design of free-form, fully three-dimensional electromagnetic scatterers, an area that has so far eluded the application of deep learning techniques.
Complex nanophotonic structures hold the potential to deliver exquisitely tailored optical responses for a range of applications. Metal-insulator-metal (MIM) metasurfaces arranged in supercells, for instance, can be tailored by geometry and material choice to exhibit a variety of absorption properties and resonant wavelengths. With this flexibility, however, comes a vast space of design possibilities that classical design paradigms struggle to effectively navigate. To overcome this challenge, here we demonstrate a tandem residual network approach to efficiently generate multiplexed supercells …
McGill University（麦吉尔大学）；National Research Council Canada（加拿大国家研究委员会）
Next-generation integrated nanophotonic device designs leverage advanced optimization techniques such as inverse design and topology optimization which achieve high performance and extreme miniaturization by optimizing a massively complex design space enabled by small feature sizes. However, unless the optimization is heavily constrained, the generated small features are not reliably fabricated, leading to optical performance degradation. Even for simpler, conventional designs, fabrication-induced performance degradation still occurs. The degree of deviation from the original design not only d…
