Pluggable multitask diffractive neural networks based on cascaded metasurfaces

Optical neural networks have significant advantages in terms of power consumption,parallelism,and high computing speed,which has intrigued extensive attention in both academic and engineering communities.It has been considered as one of the powerful tools in promoting the fields of imaging processing and object recognition.However,the existing optical system architecture cannot be reconstructed to the realization of multi-functional artificial intelligence systems simultaneously.To push the development of this issue,we propose the pluggable diffractive neural networks(P-DNN),a general paradigm resorting to the cascaded metasurfaces,which can be applied to recognize various tasks by switching internal plug-ins.As the proof-of-principle,the recognition functions of six types of handwritten digits and six types of fashions are numerical simulated and experimental demonstrated at near-infrared regimes.Encouragingly,the proposed paradigm not only improves the flexibility of the optical neural networks but paves the new route for achieving high-speed,low-power and versatile artificial intelligence systems.

Optical Neural Networks:Analysis and Prospects for 5G Applications

With the capacities of self-learning,acquainted capacities,high-speed looking for ideal arrangements,solid nonlin-ear fitting,and mapping self-assertively complex nonlinear relations,neural systems have made incredible advances and accomplished broad application over the final half-century.As one of the foremost conspicuous methods for fake insights,neural systems are growing toward high computational speed and moo control utilization.Due to the inborn impediments of electronic gadgets,it may be troublesome for electronic-implemented neural systems to make the strides these two exhibitions encourage.Optical neural systems can combine optoelectronic procedures and neural organization models to provide ways to break the bottleneck.This paper outlines optical neural networks of feedforward repetitive and spiking models to give a clearer picture of history,wildernesses,and future optical neural systems.The framework demonstrates neural systems in optic communication with the serial and parallel setup.The graphene-based laser structure for fiber optic communication is discussed.The comparison of different balance plans for photonic neural systems is made within the setting of hereditary calculation and molecule swarm optimization.In expansion,the execution comparison of routine photonic neural,time-domain with and without extending commotion is additionally expounded.The challenges and future patterns of optical neural systems on the growing scale and applications of in situ preparing nonlinear computing will hence be uncovered.

Pattern recognition in multi-synaptic photonic spiking neural networks based on a DFB-SA chip

Spiking neural networks(SNNs)utilize brain-like spatiotemporal spike encoding for simulating brain functions.Photonic SNN offers an ultrahigh speed and power efficiency platform for implementing high-performance neuromorphic computing.Here,we proposed a multi-synaptic photonic SNN,combining the modified remote supervised learning with delayweight co-training to achieve pattern classification.The impact of multi-synaptic connections and the robustness of the network were investigated through numerical simulations.In addition,the collaborative computing of algorithm and hardware was demonstrated based on a fabricated integrated distributed feedback laser with a saturable absorber(DFB-SA),where 10 different noisy digital patterns were successfully classified.A functional photonic SNN that far exceeds the scale limit of hardware integration was achieved based on time-division multiplexing,demonstrating the capability of hardware-algorithm co-computation.

Application of Neural Network in Fault Location of Optical Transport Network

Due to the increasing variety of information and services carried by optical networks, the survivability of network becomes an important problem in current research. The fault location of OTN is of great significance for studying the survivability of optical networks. Firstly, a three-channel network model is established and analyzing common alarm data, the fault monitoring points and common fault points are carried out. The artificial neural network is introduced into the fault location field of OTN and it is used to judge whether the possible fault point exists or not. But one of the obvious limitations of general neural networks is that they receive a fixedsize vector as input and produce a fixed-size vector as the output. Not only that, these models is even fixed for mapping operations (for example, the number of layers in the model). The difference between the recurrent neural network and general neural networks is that it can operate on the sequence. In spite of the fact that the gradient disappears and the gradient explodes still exist in the neural network, the method of gradient shearing or weight regularization is adopted to solve this problem, and choose the LSTM (long-short term memory networks) to locate the fault. The output uses the concept of membership degree of fuzzy theory to express the possible fault point with the probability from 0 to 1. Priority is given to the treatment of fault points with high probability. The concept of F-Measure is also introduced, and the positioning effect is measured by using location time, MSE and F-Measure. The experiment shows that both LSTM and BP neural network can locate the fault of optical transport network well, but the overall effect of LSTM is better. The localization time of LSTM is shorter than that of BP neural network, and the F1-score of LSTM can reach 0.961566888396156 after 45 iterations, which meets the accuracy and real-time requirements of fault location. Therefore, it has good application prospect and practical value to introduce neural network into the fault location field of optical transport network.

Chip-Based High-Dimensional Optical Neural Network

Parallel multi-thread processing in advanced intelligent processors is the core to realize high-speed and high-capacity signal processing systems.Optical neural network(ONN)has the native advantages of high parallelization,large bandwidth,and low power consumption to meet the demand of big data.Here,we demonstrate the dual-layer ONN with Mach-Zehnder interferometer(MZI)network and nonlinear layer,while the nonlinear activation function is achieved by optical-electronic signal conversion.Two frequency components from the microcomb source carrying digit datasets are simultaneously imposed and intelligently recognized through the ONN.We successfully achieve the digit classification of different frequency components by demultiplexing the output signal and testing power distribution.Efficient parallelization feasibility with wavelength division multiplexing is demonstrated in our high-dimensional ONN.This work provides a high-performance architecture for future parallel high-capacity optical analog computing.

Optical diagnosis of colorectal polyps using convolutional neural networks

Colonoscopy remains the gold standard investigation for colorectal cancer screening as it offers the opportunity to both detect and resect pre-malignant and neoplastic polyps.Although technologies for image-enhanced endoscopy are widely available,optical diagnosis has not been incorporated into routine clinical practice,mainly due to significant inter-operator variability.In recent years,there has been a growing number of studies demonstrating the potential of convolutional neural networks(CNN)to enhance optical diagnosis of polyps.Data suggest that the use of CNNs might mitigate the inter-operator variability amongst endoscopists,potentially enabling a“resect and discard”or“leave in”strategy to be adopted in real-time.This would have significant financial benefits for healthcare systems,avoid unnecessary polypectomies of non-neoplastic polyps and improve the efficiency of colonoscopy.Here,we review advances in CNN for the optical diagnosis of colorectal polyps,current limitations and future directions.

NEAR-INFRARED OPTICAL TOMOGRAPHY IMAGE RECONSTRUCTION APPROACH BASED ON TWO-LAYERED BP NEURAL NETWORK

An image-reconstruction approach for optical tomography is presented,in which a two-layered BP neural network is used to distinguish the tumor location.The inverse problem is solved as optimization problem by Femlab software and Levenberg–Marquardt algorithm.The concept of the average optical coefficient is proposed in this paper,which is helpful to understand the distribution of the scattering photon from tumor.The reconstructive¯µ
s by the trained network is reasonable for showing the changes of photon number transporting inside tumor tissue.It realized the fast reconstruction of tissue optical properties and provided optical OT with a new method.

Convolutional neural network for transient grating frequency-resolved optical gating trace retrieval and its algorithm optimization

A convolutional neural network is employed to retrieve the time-domain envelop and phase of few-cycle femtosecond pulses from transient-grating frequency-resolved optical gating(TG-FROG) traces.We use theoretically generated TGFROG traces to complete supervised trainings of the convolutional neural networks,then use similarly generated traces not included in the training dataset to test how well the networks are trained.Accurate retrieval of such traces by the neural network is realized.In our case,we find that networks with exponential linear unit(ELU) activation function perform better than those with leaky rectified linear unit(LRELU) and scaled exponential linear unit(SELU).Finally,the issues that need to be addressed for the retrieval of experimental data by this method are discussed.

Efficient stochastic parallel gradient descent training for on-chip optical processor

In recent years,space-division multiplexing(SDM)technology,which involves transmitting data information on multiple parallel channels for efficient capacity scaling,has been widely used in fiber and free-space optical communication sys-tems.To enable flexible data management and cope with the mixing between different channels,the integrated reconfig-urable optical processor is used for optical switching and mitigating the channel crosstalk.However,efficient online train-ing becomes intricate and challenging,particularly when dealing with a significant number of channels.Here we use the stochastic parallel gradient descent(SPGD)algorithm to configure the integrated optical processor,which has less com-putation than the traditional gradient descent(GD)algorithm.We design and fabricate a 6×6 on-chip optical processor on silicon platform to implement optical switching and descrambling assisted by the online training with the SPDG algorithm.Moreover,we apply the on-chip processor configured by the SPGD algorithm to optical communications for optical switching and efficiently mitigating the channel crosstalk in SDM systems.In comparison with the traditional GD al-gorithm,it is found that the SPGD algorithm features better performance especially when the scale of matrix is large,which means it has the potential to optimize large-scale optical matrix computation acceleration chips.

Application of the Spectrum Peak Positioning Technology Based on BP Neural Network in Demodulation of Cavity Length of EFPI Fiber Optical Sensor

An Extrinsic Fabry-Perot Interferometric (EFPI) fiber optical sensor system is an online testing system for the gas density. The system achieves the measurement of gas density information mainly by demodulating the cavity length of EF- PI fiber optical sensor. There are many ways to achieve the demodulation of the cavity length. For shortcomings of the big intensity demodulation error and complex structure of phase demodulation, this paper proposes that BP neural net-work is used to locate the special peak points in normalized interference spectrum and combining the advantages of the unimodal and bimodal measurement achieves the demodulation of the cavity length. Through online simulation and actual measurement, the results show that the peak positioning technology based on BP neural network can not only achieve high-precision demodulation of the cavity length, but also achieve an absolute measurement of cavity length in large dynamic range.

Hybrid artificial neural networks and analytical model for prediction of optical constants and bandgap energy of 3D nanonetwork silicon structures

The aim of this study is to develop a reliable method to determine optical constants for 3D-nanonetwork Si thin films manufactured using a pulsed-laser ablation technique that can be applied to other materials synthesized by this tech-nique.An analytical method was introduced to calculate optical constants from reflectance and transmittance spectra.Optical band gaps for this novel material and other important insights on the physical properties were derived from the optical constants.The existing optimization methods described in the literature were found to be complex and prone to errors while determining optical constants of opaque materials where only reflectance data is available.A supervised Deep Learning Algorithm was developed to accurately predict optical constants from the reflectance spectrum alone.The hybrid method introduced in this study was proved to be effective with an accuracy of 95%.

A review:Photonics devices,architectures,and algorithms for optical neural computing

The explosive growth of data and information has motivated various emerging non-von Neumann computational approaches in the More-than-Moore era.Photonics neuromorphic computing has attracted lots of attention due to the fascinating advantages such as high speed,wide bandwidth,and massive parallelism.Here,we offer a review on the optical neural computing in our research groups at the device and system levels.The photonics neuron and photonics synapse plasticity are presented.In addition,we introduce several optical neural computing architectures and algorithms including photonic spiking neural network,photonic convolutional neural network,photonic matrix computation,photonic reservoir computing,and photonic reinforcement learning.Finally,we summarize the major challenges faced by photonic neuromorphic computing,and propose promising solutions and perspectives.

Group Contribution Method Supervised Neural Network for Precise Design of Organic Nonlinear Optical Materials

To rationalize the design of D-π-A type organic small-molecule nonlinear optical materials,a theory guided machine learning framework is constructed.Such an approach is based on the recognition that the optical property of the molecule is predictable upon accumulating the contribution of each component,which is in line with the concept of group contribution method in thermodynamics.To realize this,a Lewis-mode group contribution method(LGC)has been developed in this work,which is combined with the multistage Bayesian neural network and the evolutionary algorithm to constitute an interactive framework(LGC-msBNN-EA).Thus,different optical properties of molecules are afforded accurately and efficientlyby using only a small data set for training.Moreover,by employing the EA model designed specifically for LGC,structural search is well achievable.The origins of the satisfying performance of the framework are discussed in detail.Considering that such a framework combines chemical principles and data-driven tools,most likely,it will be proven to be rational and efficient to complete mission regarding structure design in related fields.

Diffractive Deep Neural Networks at Visible Wavelengths

Optical deep learning based on diffractive optical elements offers unique advantages for parallel processing,computational speed,and power efficiency.One landmark method is the diffractive deep neural network(D^(2) NN)based on three-dimensional printing technology operated in the terahertz spectral range.Since the terahertz bandwidth involves limited interparticle coupling and material losses,this paper extends D^(2) NN to visible wavelengths.A general theory including a revised formula is proposed to solve any contradictions between wavelength,neuron size,and fabrication limitations.A novel visible light D^(2) NN classifier is used to recognize unchanged targets(handwritten digits ranging from 0 to 9)and targets that have been changed(i.e.,targets that have been covered or altered)at a visible wavelength of 632.8 nm.The obtained experimental classification accuracy(84%)and numerical classification accuracy(91.57%)quantify the match between the theoretical design and fabricated system performance.The presented framework can be used to apply a D^(2) NN to various practical applications and design other new applications.

Automatic Calcified Plaques Detection in the OCT Pullbacks Using Convolutional Neural Networks

Background Coronary artery calcification is a well-known marker of atherosclerotic plaque burden.High-resolution intravascular optical coherence tomography(OCT)imaging has shown the potential to characterize the details of coronary calcification in vivo.In routine clinical practice,it is a time-consuming and laborious task for clinicians to review the over 250 images in a single pullback.Besides,the imbalance label distribution within the entire pullbacks is another problem,which could lead to the failure of the classifier model.Given the success of deep learning methods with other imaging modalities,a thorough understanding of calcified plaque detection using Convolutional Neural Networks(CNNs)within pullbacks for future clinical decision was required.Methods All 33 IVOCT clinical pullbacks of 33 patients were taken from Affiliated Drum Tower Hospital,Nanjing University between December 2017 and December 2018.For ground-truth annotation,three trained experts determined the type of plaque that was present in a B-Scan.The experts assigned the labels'no calcified plaque','calcified plaque'for each OCT image.All experts were provided the all images for labeling.The final label was determined based on consensus between the experts,different opinions on the plaque type were resolved by asking the experts for a repetition of their evaluation.Before the implement of algorithm,all OCT images was resized to a resolution of 300×300,which matched the range used with standard architectures in the natural image domain.In the study,we randomly selected 26 pullbacks for training,the remaining data were testing.While,imbalance label distribution within entire pullbacks was great challenge for various CNNs architecture.In order to resolve the problem,we designed the following experiment.First,we fine-tuned twenty different CNNs architecture,including customize CNN architectures and pretrained CNN architectures.Considering the nature of OCT images,customize CNN architectures were designed that the layers were fewer than 25 layers.Then,three with good performance were selected and further deep fine-tuned to train three different models.The difference of CNNs was mainly in the model architecture,such as depth-based residual networks,width-based inception networks.Finally,the three CNN models were used to majority voting,the predicted labels were from the most voting.Areas under the receiver operating characteristic curve(ROC AUC)were used as the evaluation metric for the imbalance label distribution.Results The imbalance label distribution within pullbacks affected both convergence during the training phase and generalization of a CNN model.Different labels of OCT images could be classified with excellent performance by fine tuning parameters of CNN architectures.Overall,we find that our final result performed best with an accuracy of 90%of'calcified plaque'class,which the numbers were less than'no calcified plaque'class in one pullback.Conclusions The obtained results showed that the method is fast and effective to classify calcific plaques with imbalance label distribution in each pullback.The results suggest that the proposed method could be facilitating our understanding of coronary artery calcification in the process of atherosclerosis andhelping guide complex interventional strategies in coronary arteries with superficial calcification.

Combined Transmission Interference Spectrum of No Core Fiber and BP Neural Network for Concentration Sensing Research

To investigate wavelength response of the no core fiber（NCF）interference spectrum to concentration,a three-layer back propagation（BP）neural network model was established to optimize the concentration sensing data.In this method,the measured wavelength and the corresponding concentration were trained by a BP neural network,so that the accuracy of the measurement system was optimized.The wavelength was used as the training set and got into the input layer of the three layer BP network model which is used as the input value of the network,and the corresponding actual concentration value was used as the output value of the network,and the optimal network structure was trained.This paper discovers a preferable correlation between the predicted value and the actual value,where the former is approximately equal to the latter.The correlation coefficients of the measured and predicted values for a sucrose concentration were 1.000 89 and 1.003 94;similarly,correlations of0.999 51 and 1.018 8 for a glucose concentration were observed.The results demonstrate that the BP neural network can improve the prediction accuracy of the nonlinear relationship between the interference spectral data and the concentration in NCF sensing systems.

Artifcial eur ral networks based estimation of optical parameters by diffuse reflectance imaging under in vitro conditions

Optical parameters(properties)of tissue-mimicking phantoms are determined through nonin-vasive optical imaging.Objective of this study is to decompose obtained difuse reflectance into these optical properties such as absorption and scattering coefficients.To do so,transmission spectroscopy is firstly used to measure the coefficients via an experimental setup.Next,the optical properties of each characterized phantom are input for Monte Carlo(MC)simulations to get diffuse reflectance.Also,a surface image for each single phantom with its known optical properties is obliquely captured due to reflectance-based geometrical setup using CMOS camera that is positioned at 5°angle to the phantoms.For the illumination of light,a laser light source at 633 nm wavelength is preferred,because optical properties of different components in a biological tissue on that wavelength are nonoverlapped.During in vitro measurements,we prepared 30 different mixture samples adding clinoleic intravenous lipid emulsion(CILE)and evans blue(EB)dye into a distilled water.Finally,all obtained difuse reflectance values are used to estimate the optical coefficients by artificial neural networks(ANNs)in inverse modeling.For a biological tissue it is found that the simulated and measured values in our results are in good agreement.

Enhanced Detection of Glaucoma on Ensemble Convolutional Neural Network for Clinical Informatics

Irretrievable loss of vision is the predominant result of Glaucoma in the retina.Recently,multiple approaches have paid attention to the automatic detection of glaucoma on fundus images.Due to the interlace of blood vessels and the herculean task involved in glaucoma detection,the exactly affected site of the optic disc of whether small or big size cup,is deemed challenging.Spatially Based Ellipse Fitting Curve Model(SBEFCM)classification is suggested based on the Ensemble for a reliable diagnosis of Glaucomain theOptic Cup(OC)and Optic Disc(OD)boundary correspondingly.This research deploys the Ensemble Convolutional Neural Network(CNN)classification for classifying Glaucoma or Diabetes Retinopathy(DR).The detection of the boundary between the OC and the OD is performed by the SBEFCM,which is the latest weighted ellipse fitting model.The SBEFCM that enhances and widens the multi-ellipse fitting technique is proposed here.There is a preprocessing of input fundus image besides segmentation of blood vessels to avoid interlacing surrounding tissues and blood vessels.The ascertaining of OCandODboundary,which characterizedmany output factors for glaucoma detection,has been developed by EnsembleCNNclassification,which includes detecting sensitivity,specificity,precision,andArea Under the receiver operating characteristic Curve(AUC)values accurately by an innovative SBEFCM.In terms of contrast,the proposed Ensemble CNNsignificantly outperformed the current methods.

A CASCADED MODEL OF NEURAL NETWORK FOR PATTERN RECOGNITION

A cascaded model of neural network and its learning algorithm suitable for opticalimplementation are proposed.Computer simulations have shown that this model may successfullybe applied to an error-tolerance pattern recognitions of multiple 3-D targets with arbitrary spatialorientations.

Hamilton Number Neural Network Model

The stability and the sequential dynamics of the sixteen-state Hamiltnn neural network model have been discussed with the energy function. The optical implementation scheme about Hamilton number vector-matrix multiplication is simply discussed.