Advancing statistical learning and artificial intelligence in nanophotonics inverse design

Wang, Qizhou; Makarenko, Maksim; Burguete-Lopez, Arturo; Getman, F.; Fratalocchi, Andrea

doi:10.1515/nanoph-2021-0660

Cited by 28 publications

(18 citation statements)

References 116 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We will not review here in depth the different types of machine learning algorithms applied to photonic design, due to the vast and rapidly developing field, but we will introduce the concepts related to this new frontier in computer science applied to metasurface design. For more in-depth discussions, we refer the reviewer to recent reviews on the topic. − …”

Section: Fundamentalsmentioning

confidence: 99%

Optical Metasurfaces for Energy Conversion

et al. 2022

View full text Add to dashboard Cite

Nanostructured surfaces with designed optical functionalities, such as metasurfaces, allow efficient harvesting of light at the nanoscale, enhancing light–matter interactions for a wide variety of material combinations. Exploiting light-driven matter excitations in these artificial materials opens up a new dimension in the conversion and management of energy at the nanoscale. In this review, we outline the impact, opportunities, applications, and challenges of optical metasurfaces in converting the energy of incoming photons into frequency-shifted photons, phonons, and energetic charge carriers. A myriad of opportunities await for the utilization of the converted energy. Here we cover the most pertinent aspects from a fundamental nanoscopic viewpoint all the way to applications.

show abstract

Section: Fundamentalsmentioning

confidence: 99%

Optical Metasurfaces for Energy Conversion

et al. 2022

View full text Add to dashboard Cite

show abstract

“…metasurface technology. Modern metasurface design approaches [51,52] usually rely on a library of pre-computed metasurface responses and polynomial fitting to further generalize the relationship between design parameters and the device performance. We, instead, design our metasurface optical projectors via ALFRED [20], a hybrid inverse design scheme that combines classical optimization and deep learning [52].…”

Section: Reconstruction By Sparse Coding and Deep Learningmentioning

confidence: 99%

“…Modern metasurface design approaches [51,52] usually rely on a library of pre-computed metasurface responses and polynomial fitting to further generalize the relationship between design parameters and the device performance. We, instead, design our metasurface optical projectors via ALFRED [20], a hybrid inverse design scheme that combines classical optimization and deep learning [52]. In this work, we significantly extend the capabilities of the original code by adding differentiability, physical-model regularization, and complex decoder projectors able to tackle different computer vision tasks and perform thousands of parameter optimizations through the supervised end-to-end learning process.…”

Section: Reconstruction By Sparse Coding and Deep Learningmentioning

confidence: 99%

Real-time Hyperspectral Imaging in Hardware via Trained Metasurface Encoders

Makarenko¹,

Burguete-Lopez²,

Wang³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Hyperspectral imaging has attracted significant attention to identify spectral signatures for image classification and automated pattern recognition in computer vision. State-ofthe-art implementations of snapshot hyperspectral imaging rely on bulky, non-integrated, and expensive optical elements, including lenses, spectrometers, and filters. These macroscopic components do not allow fast data processing for, e.g. real-time and high-resolution videos. This work introduces Hyplex™, a new integrated architecture addressing the limitations discussed above. Hyplex™ is a CMOS-compatible, fast hyperspectral camera that replaces bulk optics with nanoscale metasurfaces inversely designed through artificial intelligence. Hyplex™ does not require spectrometers but makes use of conventional monochrome cameras, opening up the possibility for real-time and high-resolution hyperspectral imaging at inexpensive costs. Hyplex™ exploits a modeldriven optimization, which connects the physical metasurfaces layer with modern visual computing approaches based on end-to-end training. We design and implement a prototype version of Hyplex™ and compare its performance against the state-of-the-art for typical imaging tasks such as spectral reconstruction and semantic segmentation. In all benchmarks, Hyplex™ reports the smallest reconstruction error. We additionally present what is, to the best of our knowledge, the largest publicly available labeled hyperspectral dataset for semantic segmentation. 1

show abstract

“…Recently, neural networks (NNs) have shown great potential in photonic signal processing in various areas [8][9][10][11][12][13][14]. Inverse design using a NN where a trained model is built to predict inverse geometric solutions for desired spectral transmission, has also emerged as a powerful tool for the rapid design of ultra-compact photonic integrated devices with various functionalities [6,7,[15][16][17][18][19][20][21][22][23][24][25][26][27][28]. However, due to the non-unique relationship between the geometry and the spectral transmission, the NN-based inverse design for nano-photonic devices is usually hard to converge [27,28].…”

Section: Introductionmentioning

confidence: 99%

Inverse design of a nano-photonic wavelength demultiplexer with a deep neural network approach

Yuan

Song

et al. 2022

Opt. Express

View full text Add to dashboard Cite

In this paper, we propose a pre-trained-combined neural network (PTCN) as a comprehensive solution to the inverse design of an integrated photonic circuit. By utilizing both the initially pre-trained inverse and forward model with a joint training process, our PTCN model shows remarkable tolerance to the quantity and quality of the training data. As a proof of concept demonstration, the inverse design of a wavelength demultiplexer is used to verify the effectiveness of the PTCN model. The correlation coefficient of the prediction by the presented PTCN model remains greater than 0.974 even when the size of training data is decreased to 17%. The experimental results show a good agreement with predictions, and demonstrate a wavelength demultiplexer with an ultra-compact footprint of 2.6×2.6μm 2 , a high transmission efficiency with a transmission loss of -2dB, a low reflection of -10dB, and low crosstalk around -7dB simultaneously.

show abstract

Advancing statistical learning and artificial intelligence in nanophotonics inverse design

Cited by 28 publications

References 116 publications

Optical Metasurfaces for Energy Conversion

Optical Metasurfaces for Energy Conversion

Real-time Hyperspectral Imaging in Hardware via Trained Metasurface Encoders

Inverse design of a nano-photonic wavelength demultiplexer with a deep neural network approach

Contact Info

Product

Resources

About