Matrix computation, as a fundamental building block of information processing in science and technology, contributes most of the computational overheads in modern signal processing and artificial intelligence algorithms. Photonic accelerators are designed to accelerate specific categories of computing in the optical domain, especially matrix multiplication, to address the growing demand for computing resources and capacity. Photonic matrix multiplication has much potential to expand the domain of telecommunication, and artificial intelligence benefiting from its superior performance. Recent research in photonic matrix multiplication has flourished and may provide opportunities to develop applications that are unachievable at present by conventional electronic processors. In this review, we first introduce the methods of photonic matrix multiplication, mainly including the plane light conversion method, Mach-Zehnder interferometer method and wavelength division multiplexing method. We also summarize the developmental milestones of photonic matrix multiplication and the related applications. Then, we review their detailed advances in applications to optical signal processing and artificial neural networks in recent years. Finally, we comment on the challenges and perspectives of photonic matrix multiplication and photonic acceleration. 

/pdf/photonic-matrix-multiplication-lights-up-photonic-3ufkf760.pdf

Photonic matrix multiplication lights up photonic accelerator and beyond

/pdf/photonic-matrix-multiplication-lights-up-photonic-fmtloxf9.pdf

Invisibility has been a long-standing topic, owing to its crucial importance in both academia and industry. At the close of the last century, the advent of metamaterials unprecedentedly ignited the enthusiasm toward pursuing invisibility cloaks, but despite two decades of research, it has yet to mature into a practical method. In this Perspective, we first present a review of conventional and state-of-the-art invisibility methodologies, as well as their intrinsic challenges. To take a further step, we focus on the conjunction of two burgeoning disciplines—metamaterials and deep learning—to facilitate an intelligent self-adaptive invisibility cloak for dynamic surroundings. Finally, future directions are described, which include possible development directions and other practice-oriented metadevices driven by deep learning.

A perspective on the next generation of invisibility cloaks—Intelligent cloaks

Abstract With the growing interest in the field of artificial materials, more advanced and sophisticated functionalities are required from phononic crystals and acoustic metamaterials. This implies a high computational effort and cost, and still the efficiency of the designs may be not sufficient. With the help of third-wave artificial intelligence technologies, the design schemes of these materials are undergoing a new revolution. As an important branch of artificial intelligence, machine learning paves the way to new technological innovations by stimulating the exploration of structural design. Machine learning provides a powerful means of achieving an efficient and accurate design process by exploring nonlinear physical patterns in high-dimensional space, based on data sets of candidate structures. Many advanced machine learning algorithms, such as deep neural networks, unsupervised manifold clustering, reinforcement learning and so forth, have been widely and deeply investigated for structural design. In this review, we summarize the recent works on the combination of phononic metamaterials and machine learning. We provide an overview of machine learning on structural design. Then discuss machine learning driven on-demand design of phononic metamaterials for acoustic and elastic waves functions, topological phases and atomic-scale phonon properties. Finally, we summarize the current state of the art and provide a prospective of the future development directions.

Intelligent on-demand design of phononic metamaterials

In the recent years, we observe a dramatic boost of research in photonics empowered by the concepts of machine learning and artificial intelligence. The corresponding photonic systems, to which this new methodology is applied, can range from traditional optical waveguides to nanoantennas and metasurfaces, and these novel approaches underpin the fundamental principles of light-matter interaction developed for a smart design of intelligent photonic devices. Concepts and approaches of artificial intelligence and machine learning penetrate rapidly into the fundamental physics of light, and they provide effective tools for the study of the field of metaphotonics driven by optically-induced electric and magnetic resonances. Here, we introduce this new field with its application to metaphotonics and also present a summary of the basic concepts of machine learning with some specific examples developed and demonstrated for metasystems and metasurfaces. 

Intelligent metaphotonics empowered by machine learning

Being invisible at will has been a long-standing dream for centuries, epitomized by numerous legends; humans have never stopped their exploration steps to realize this dream. Recent years have witnessed a breakthrough in this search due to the advent of transformation optics, metamaterials, and metasurfaces. However, the previous metasurface cloaks typically work in a reflection manner that relies on a high-reflection background, thus limiting the applications. Here, we propose an easy yet viable approach to realize the transmitted metasurface cloak, just composed of two planar metasurfaces to hide an object inside, such as a cat. To tackle the hard-to-converge issue caused by the nonuniqueness phenomenon, we deploy a tandem neural network (T-NN) to efficiently streamline the inverse design. Once pretrained, the T-NN can work for a customer-desired electromagnetic response in one single forward computation, saving a great amount of time. Our work opens a new avenue to realize a transparent invisibility cloak, and the tandem-NN can also inspire the inverse design of other metamaterials and photonics.

/pdf/realizing-transmitted-metasurface-cloak-by-a-tandem-neural-4aogjtyll0.pdf

Realizing transmitted metasurface cloak by a tandem neural network

The embodiment of the invention provides an electromagnetic metasurface design method and device based on deep learning. The electromagnetic metasurface design method comprises the steps: carrying outforward prediction of electromagnetic metasurface design according to a deep learning model M1 to obtain a forward mapping relation between electromagnetic metasurface structures/arrangements and corresponding optical responses; and according to the deep learning model M2, carrying out reverse prediction of the electromagnetic super-design to obtain a reverse mapping relationship between the electromagnetic super-surface structure/arrangement and the corresponding optical response. In forward prediction, the defect that a traditional method depends on numerical simulation iteration to solve the problem that a Maxwell equation is complex and time-consuming is overcome, the calculation complexity is greatly reduced, and the calculation time is shortened. In the reverse design, the geometricparameters of the electromagnetic super-unit are effectively discovered and optimized by using the reversely designed deep learning model obtained by training, and the electromagnetic super-surface structure/arrangement under different incident wave information is obtained in real time, so that the electromagnetic super-surface design which is customized by a user and can be strained as requiredis realized.

Jia Yuetian

Papers

Realizing transmitted metasurface cloak by a tandem neural network

Electromagnetic metasurface design method and device based on deep learning