检索结果-内蒙古大学图书馆

Applying machine learning to optical metrology: a review

MEASUREMENT SCIENCE AND TECHNOLOGY 2025年第1期36卷 012002-012002页

作者： Xue, Ruidong Hooshmand, Helia Isa, Mohammed Piano, Samanta Leach, Richard Univ Nottingham Fac Engn Mfg Metrol Team Nottingham England

This literature review investigates the integration of machine learning (ML) into optical metrology, unveiling enhancements in both efficiency and effectiveness of measurement processes. With a focus on phase demodulation, unwrapping, and phase-to-height conversion, the review highlights how ML algorithms have transformed traditional optical metrology techniques, offering improved speed, accuracy, and data processing capabilities. Efficiency improvements are underscored by advancements in data generation, intelligent sampling, and processing strategies, where ML algorithms have accelerated the metrological evaluations. Effectiveness is enhanced in measurement precision, with ML providing robust solutions to complex pattern recognition and noise reduction challenges. Additionally, the role of parallel computing using graphics processing units and field programmable gate arrays is emphasised, showcasing their importance in supporting the computationally intensive ML algorithms for real-time processing. This review culminates in identifying future research directions, emphasising the potential of advanced ML models and broader applications within optical metrology. Through this investigation, the review articulates a future where optical metrology, empowered by ML, achieves improved levels of operational efficiency and effectiveness.

关键词： machine learning artificial intelligence neural networks optical metrology

来源：评论

学校读者我要写书评

暂无评论

VCSEL-based photonic spiking neural networks for ultrafast detection and tracking

引用

NEUROMORPHIC computing AND ENGINEERING 2024年第1期4卷 014010-014010页

作者： Robertson, Joshua Kirkland, Paul Di Caterina, Gaetano Hurtado, Antonio Univ Strathclyde Inst Photon SUPA Dept Phys Glasgow Scotland Univ Strathclyde Dept Elect & Elect Engn Glasgow Scotland

Inspired by efficient biological spike-based neural networks, we demonstrate for the first time the detection and tracking of target patterns in image and video inputs at high-speed rates with networks of multiple artificial spiking optical neurons. Using photonic systems of in-parallel spiking vertical cavity surface emitting lasers (VCSELs), we demonstrate the implementation of multiple convolutional kernel operators which, in combination with optical spike signalling, enable the detection and tracking of target features in images/video feeds at an ultrafast photonic operation speed of 1 ns per pixel. Alongside a single layer optical spiking neural network (SNN) demonstration, a multi-layer network of photonic (GHz-rate) spike-firing neurons is reported where the photonic system successfully tracks a large complex feature (Handwritten Digit 3). The consecutive photonic layers perform spike-enabled image reduction and convolution operations, and interact with a software-implemented SNN, that learns the feature patterns that best identify the target to provide a high detection efficiency even in the presence of a distractor feature. This work therefore highlights the effectiveness of combining neuromorphic photonic hardware and software SNNs, for efficient learning and ultrafast operation, thanks to the use of spiking light signals, towards tackling complex AI and computer vision problems.

关键词： neuromorphic computing semiconductor laser spiking neural networks image processing detection and tracking

来源：评论

学校读者我要写书评

暂无评论

Annealing-inspired training of an optical neural network with ternary weights

引用

COMMUNICATIONS PHYSICS 2025年第1期8卷 1-10页

作者： Skalli, Anas Goldmann, Mirko Haghighi, Nasibeh Reitzenstein, Stephan Lott, James A. Brunner, Daniel Univ Marie & Louis Pasteur CNRS Inst FEMTO ST UMR 6174 Besancon France Tech Univ Berlin Berlin Germany

Artificial neural networks (ANNs) represent a fundamentally connectionist and distributed approach to computing, and as such they differ from classical computers that utilize the von Neumann architecture. This has revived research interest in new unconventional hardware for more efficient ANNs rather than emulating them on traditional machines. To fully leverage ANNs, optimization algorithms must account for hardware limitations and imperfections. Photonics offers a promising platform with scalability, speed, energy efficiency, and parallel processing capabilities. However, fully autonomous optical neural networks (ONNs) with in-situ learning are scarce. In this work, we propose and demonstrate a ternary weight high-dimensional semiconductor laser-based ONN and introduce a method for achieving ternary weights using Boolean hardware, enhancing the ONN's information processing capabilities. Furthermore, we design an in-situ optimization algorithm that is compatible with both Boolean and ternary weights. Our algorithm results in benefits, both in terms of convergence speed and performance. Our experimental results show the ONN's long-term inference stability, with a consistency above 99% for over 10 h. Our work is of particular relevance in the context of in-situ learning under restricted hardware resources, especially since minimizing the power consumption of auxiliary hardware is crucial to preserving efficiency gains achieved by non-von Neumann ANN implementations.

关键词： In situ processing

来源：评论

学校读者我要写书评

暂无评论

Inverse Design of Unitary Transmission Matrices in Silicon Photonic Coupled Waveguide Arrays Using a neural Adjoint Model

引用

ACS PHOTONICS 2025年第3期12卷 1480-1493页

作者： Radford, Thomas W. Wiecha, Peter R. Politi, Alberto Zeimpekis, Ioannis Muskens, Otto L. Univ Southampton Sch Phys & Astron Southampton SO17 1BJ England Univ Toulouse CNRS LAAS F-31031 Toulouse France Univ Southampton Sch Elect & Comp Sci Southampton SO17 1BJ England Univ Southampton Optoelect Res Ctr Southampton SO17 1BJ England

The development of low-loss reconfigurable integrated optical devices enables further research into technologies including photonic signal processing, analogue quantum computing, and optical neural networks. Here, we introduce digital patterning of coupled waveguide arrays as a platform capable of implementing unitary matrix operations. Determining the required device geometry for a specific optical output is computationally challenging and requires a robust and versatile inverse design protocol. In this work we present an approach using high speed neural network surrogate-based gradient optimization, capable of predicting patterns of refractive index perturbations based on switching of the ultralow loss chalcogenide phase change material, antimony triselinide (Sb2Se3). Results for a 3 x 3 silicon waveguide array are presented, demonstrating control of both amplitude and phase for each transmission matrix element. Network performance is studied using neural network optimization tools such as data set augmentation and supplementation with random noise, resulting in an average fidelity of 0.94 for unitary matrix targets. Our results show that coupled waveguide arrays with perturbation patterns offer new routes for achieving programmable unitary operators, or Hamiltonians for quantum simulators, with a reduced footprint compared to conventional interferometer-mesh technology.

关键词： inverse design deep learning neuraladjoint programmable photonic devices phase changematerials Sb2Se3

来源：评论

学校读者我要写书评

暂无评论

A Comparison Between ANN and RBFNN for Signal Demodulation in Photon Counting based optical Wireless Communications 12

A Comparison Between ANN and RBFNN for Signal Demodulation i...

引用

12th International Conference on Intelligent computing and Wireless optical Communications (ICWOC)

作者： He, Cuiwei Chen, Chen Japan Adv Inst Sci & Technol Sch Informat Sci Nomi Ishikawa Japan Chongqing Univ Sch Microelect & Commun Engn Chongqing Peoples R China

ISBN: (纸本)9798350374377

The use of silicon photomultiplier (SiPM) sensors in optical wireless communication (OWC) has shown very promising performance when detecting optical signals with low intensity. However, the recovery period of individual SiPM microcells introduces a nonlinear response and also causes intersymbol interference (ISI) when the transmission data rate is high. The use of a neural network based demodulator has been demonstrated to be more accurate for signal demodulation, especially when the SiPM is nonlinear. Among various neural network algorithms, artificial neural network (ANN) and radial basis function neural network (RBFNN) are two promising candidates. In this paper, we focus on comparing these two neural network structures for signal demodulation. First, to intuitively explain how these two neural networks can be applied for signal demodulation, we consider their structures with a small number of neurons and the constellations of binary phase-shift keying (BPSK). In particular, for ANN, we consider a single neuron so that the signal demodulation step is equivalent to a logistic regression task. In the case of RBFNN, we consider a network with six neurons. By showing the detailed updating process of both networks, we demonstrate that both ANN and RBFNN can achieve almost identical classification boundaries in the Euclidean space. Finally, we consider more complex structures for both ANN and RBFNN, along with larger constellation sizes. The results show that these two structures achieve identical performance for a wide range of irradiances.

关键词： Photon counting Artificial intelligence (AI) SiPM optical wireless communications Artificial neural network (ANN) Multi-layer perceptron (MLP) Radial basis function neural network (RBFNN)

来源：评论

学校读者我要写书评

暂无评论

Inverse design of an on-chip optical response predictor enabled by a deep neural network

引用

OPTICS EXPRESS 2023年第2期31卷 2049-2060页

作者： Kim, Junhyeong Neseli, Berkay Kim, Jae-Yong Yoon, Jinhyeong Yoon, Hyeonho Park, Hyo-Hoon Kurt, Hamza Korea Adv Inst Sci & Technol Sch Elect Engn Daejeon 34141 South Korea

We proposed inverse-designed nanophotonic waveguide devices which have the desired optical responses in the wide band of 1450-1650 nm. The proposed devices have an ultra-compact size of just 1.5 mu m x 3.0 mu m and are designed on a silicon-on-insulator (SOI) waveguide platform. Individual nano-pixels with dimensions of 150 nm x 150 nm were made of either silicon or silicon dioxide, and the materials for the 200 total cells were determined using a trained deep neural network. While training the two networks, the hyperparameter optimization method was applied to make the training process efficient. We then fabricated the proposed devices using a CMOS-compatible fabrication process, and experimentally verified the fabricated device performance.

关键词： Electron beam lithography Genetic algorithms neural networks optical computing Polarization splitters Silica

来源：评论

学校读者我要写书评

暂无评论

Fully forward mode training for optical neural networks

引用

NATURE 2024年第8024期632卷 280页

作者： Xue, Zhiwei Zhou, Tiankuang Xu, Zhihao Yu, Shaoliang Dai, Qionghai Fang, Lu Tsinghua Univ Dept Elect Engn Beijing Peoples R China Tsinghua Univ Beijing Natl Res Ctr Informat Sci & Technol Beijing Peoples R China Tsinghua Univ Inst Brain & Cognit Sci Beijing Peoples R China Tsinghua Univ Shenzhen Int Grad Sch Shenzhen Peoples R China Zhejiang Lab Res Ctr Intelligent Optoelect Comp Hangzhou Peoples R China Tsinghua Univ Dept Automat Beijing Peoples R China

optical computing promises to improve the speed and energy efficiency of machine learning applications1-6. However, current approaches to efficiently train these models are limited by in silico emulation on digital computers. Here we develop a method called fully forward mode (FFM) learning, which implements the compute-intensive training process on the physical system. The majority of the machine learning operations are thus efficiently conducted in parallel on site, alleviating numerical modelling constraints. In free-space and integrated photonics, we experimentally demonstrate optical systems with state-of-the-art performances for a given network size. FFM learning shows training the deepest optical neural networks with millions of parameters achieves accuracy equivalent to the ideal model. It supports all-optical focusing through scattering media with a resolution of the diffraction limit;it can also image in parallel the objects hidden outside the direct line of sight at over a kilohertz frame rate and can conduct all-optical processing with light intensity as weak as subphoton per pixel (5.40 x 1018- operations-per-second-per-watt energy efficiency) at room temperature. Furthermore, we prove that FFM learning can automatically search non-Hermitian exceptional points without an analytical model. FFM learning not only facilitates orders-of-magnitude-faster learning processes, but can also advance applied and theoretical fields such as deep neural networks, ultrasensitive perception and topological photonics. We present fully forward mode learning, which conducts machine learning operations on site, leading to faster learning and promoting advancement in numerous fields.

关键词： Computer science Electrical and electronic engineering

来源：评论

学校读者我要写书评

暂无评论

Photonic Spiking neural networks with Highly Efficient Training Protocols for Ultrafast Neuromorphic computing Systems

INTELLIGENT COMPUTING

引用

INTELLIGENT computing 2023年 2卷

作者： Owen-Newns, Dafydd Robertson, Joshua Hejda, Matej Hurtado, Antonio Univ Strathclyde Inst Photon SUPA Dept Phys Glasgow Scotland

Photonic technologies offer great prospects for novel, ultrafast, energy-efficient, and hardware-friendly neuromorphic (brain-like) computing platforms. Moreover, neuromorphic photonic approaches based on ubiquitous, technology-mature, and low-cost vertical-cavity surface-emitting lasers (VCSELs) (devices found in fiber-optic transmitters, mobile phones, and automotive sensors) are of particular interest. Given that VCSELs have shown the ability to realize neuronal optical spiking responses (at ultrafast GHz rates), their use in spike-based information-processing systems has been proposed. In this study, spiking neural network (SNN) operation, based on a hardware-friendly photonic system of just one VCSEL, is reported alongside a novel binary weight "significance" training scheme that fully capitalizes on the discrete nature of the optical spikes used by the SNN to process input information. The VCSEL-based photonic SNN was tested with a highly complex multivariate classification task (MADELON) before its performance was compared using a traditional least-squares training method and an alternative novel binary weighting scheme. Excellent classification accuracies of >94% were achieved by both training methods, exceeding the benchmark performance of the dataset in a fraction of the processing time. The newly reported training scheme also dramatically reduces the training set size requirements and the number of trained nodes (<= 1% of the total network node count). This VCSEL-based photonic SNN, in combination with the reported "significance" weighting scheme, therefore grants ultrafast spike-based optical processing highly reduced training requirements and hardware complexity for potential application in future neuromorphic systems and artificial intelligence applications.

关键词： Digital storage

来源：评论

学校读者我要写书评

暂无评论

Characterizing Coherent Integrated Photonic neural networks Under Imperfections

引用

JOURNAL OF LIGHTWAVE TECHNOLOGY 2023年第5期41卷 1464-1479页

作者： Banerjee, Sanmitra Nikdast, Mahdi Chakrabarty, Krishnendu Duke Univ Dept Elect & Comp Engn Durham NC 27708 USA Colorado State Univ Dept Elect & Comp Engn Ft Collins CO 80523 USA

Integrated photonic neural networks (IPNNs) are emerging as promising successors to conventional electronic AI accelerators as they offer substantial improvements in computing speed and energy efficiency. In particular, coherent IPNNs use arrays of Mach-Zehnder interferometers (MZIs) for unitary transformations to perform energy-efficient matrix-vector multiplication. However, the underlying MZI devices in IPNNs are susceptible to uncertainties stemming from optical lithographic variations and thermal crosstalk and can experience imprecisions due to non-uniform MZI insertion loss and quantization errors due to low-precision encoding in the tuned phase angles. In this article, we, for the first time, systematically characterize the impact of such uncertainties and imprecisions (together referred to as imperfections) in IPNNs using a bottom-up approach. We show that their impact on IPNN accuracy can vary widely based on the tuned parameters (e.g., phase angles) of the affected components, their physical location, and the nature and distribution of the imperfections. To improve reliability measures, we identify critical IPNN building blocks that, under imperfections, can lead to catastrophic degradation in the classification accuracy. We show that under multiple simultaneous imperfections, the IPNN inferencing accuracy can degrade by up to 46%, even when the imperfection parameters are restricted within a small range. Our results also indicate that the inferencing accuracy is sensitive to imperfections affecting the MZIs in the linear layers next to the input layer of the IPNN.

关键词： Adaptive optics Transmission line matrix methods Neurons optical waveguides optical interferometry optical crosstalk Uncertainty neural networks phase angles process variations quantization errors silicon photonics

来源：评论

学校读者我要写书评

暂无评论

Integrated input system for partially coherent light control in optical neural networks for remote sensing

引用

OPTICS AND LASER TECHNOLOGY 2025年 189卷

作者： Li, Hongmin Liu, Xun Xu, Haonan Peng, Zhenxing Zhou, Xu Li, Wei Lu, Yao Wu, Qiang Xu, Jingjun Nankai Univ TEDA Inst Appl Phys Minist Educ Key Lab Weak Light Nonlinear Photon Tianjin 300457 Peoples R China Nankai Univ Sch Phys Tianjin 300457 Peoples R China China Acad Space Technol Beijing Inst Space Mech & Elect Beijing 100094 Peoples R China

Laser-based all-optical neural networks (ONNs) offer remarkable advantages in processing speed, interference resistance, and energy efficiency, making them highly applicable for fields such as machine vision and image classification recognition. However, in practical scenarios, laser coherence can degrade due to transmission interference, posing a significant challenge to the performance and reliability of ONNs. Against this backdrop, the exploration of neural network technologies based on partially coherent light (PCL) becomes crucial, underscoring the urgent need for efficient and convenient methods to generate, calibrate, and precisely control PCL. Herein, we propose an integrated ONNs input system incorporating the functions of generation, calibration, precise control, and image reconstruction of PCL. The system not only enables the precise conversion and calibration from lasers to specific coherence levels but also achieves high-repetition-rate, high-quality image reconstruction. Consequently, it rapidly provides abundant high-quality samples support for the input layer of ONNs, significantly enhancing training efficiency. The system focuses on the future recognition and analysis of large, high-resolution images by ONNs, demonstrating robust scalability and further emphasizing the advantages of ONNs as highly efficient parallel systems. Our research effectively addresses the challenges posed by laser coherence degradation, offering new insights into the development and practical implementation of ONNs. Furthermore, it paves the way for their seamless integration into fields such as aerospace remote sensing and image classification recognition.

关键词： Light field imaging optical reconstruction Partially coherent light All-optical computing

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：