The rapid advancement of deep learning and the emergence of largescale neural models,such as bidirectional encoder representations from transformers(BERT),generative pre-trained transformer(GPT),and large language mod...The rapid advancement of deep learning and the emergence of largescale neural models,such as bidirectional encoder representations from transformers(BERT),generative pre-trained transformer(GPT),and large language model Meta AI(LLaMa),have brought significant computational and energy challenges.Neuromorphic computing presents a biologically inspired approach to addressing these issues,leveraging event-driven processing and in-memory computation for enhanced energy efficiency.This survey explores the intersection of neuromorphic computing and large-scale deep learning models,focusing on neuromorphic models,learning methods,and hardware.We highlight transferable techniques from deep learning to neuromorphic computing and examine the memoryrelated scalability limitations of current neuromorphic systems.Furthermore,we identify potential directions to enable neuromorphic systems to meet the growing demands of modern AI workloads.展开更多
Recently,due to the availability of big data and the rapid growth of computing power,artificial intelligence(AI)has regained tremendous attention and investment.Machine learning(ML)approaches have been successfully ap...Recently,due to the availability of big data and the rapid growth of computing power,artificial intelligence(AI)has regained tremendous attention and investment.Machine learning(ML)approaches have been successfully applied to solve many problems in academia and in industry.Although the explosion of big data applications is driving the development of ML,it also imposes severe challenges of data processing speed and scalability on conventional computer systems.Computing platforms that are dedicatedly designed for AI applications have been considered,ranging from a complement to von Neumann platforms to a“must-have”and stand-alone technical solution.These platforms,which belong to a larger category named“domain-specific computing,”focus on specific customization for AI.In this article,we focus on summarizing the recent advances in accelerator designs for deep neural networks(DNNs)-that is,DNN accelerators.We discuss various architectures that support DNN executions in terms of computing units,dataflow optimization,targeted network topologies,architectures on emerging technologies,and accelerators for emerging applications.We also provide our visions on the future trend of AI chip designs.展开更多
The terahertz(THz) far-field radiation properties of a butterfly-shaped photoconductive antenna(PCA) were experimentally studied using a home-built THz time-domain spectroscopy(THz-TDS) setup.To distinguish the contri...The terahertz(THz) far-field radiation properties of a butterfly-shaped photoconductive antenna(PCA) were experimentally studied using a home-built THz time-domain spectroscopy(THz-TDS) setup.To distinguish the contribution of in-gap photocurrent and antenna structure to far-field radiation,polarization-dependent THz field was measured and quantified as the illuminating laser beam moved along the bias field within the gap region of electrodes. The result suggests that, although the far-field THz radiation originates from the in-gap photocurrent, the antenna structure of butterfly-shaped PCA dominates the overall THz radiation. In addition, to explore the impact of photoconductive material,radiation properties of butterfly-shaped PCAs fabricated on both low-temperature-grown GaAs(LT-GaAs) and semi-insulating GaAs(Si-GaAs) were characterized and compared. Consistent with previous experiments, it is observed that while Si-GaAs-based PCA can emit higher THz field than LT-GaAs-based PCA at low laser power, it would saturate more severely as laser power increased and eventually be surpassed by LT-GaAs-based PCA. Beyond that, it is found the severe saturation effect of Si-GaAs was due to the longer carrier lifetime and higher carrier mobility, which was confirmed by the numerical simulation.展开更多
A dual-band and high-isolation shared-aperture antenna for unmanned aerial vehicle(UAV)platforms has been proposed.This shared-aperture antenna consists of a rectangular monopole antenna and a 4-element multiple input...A dual-band and high-isolation shared-aperture antenna for unmanned aerial vehicle(UAV)platforms has been proposed.This shared-aperture antenna consists of a rectangular monopole antenna and a 4-element multiple input multiple output(MIMO)antenna.In order to increase the isolation,several double split ring metamaterial(MTM)structures are introduced between antenna elements.The antenna radiator and the MTM structure are located on the front and back of the same dielectric substrate,respectively,and are perpendicular to a circular metal plate.The overall size of the antenna substrate is 124 mm×38 mm×1.016 mm.Moreover,the antenna prototype is constructed and measured,and the simulated and measured results are in good agreement.The measured results show that the-10 dB bandwidth of the monopole antenna is 1.92 GHz to 2.75 GHz,and the common-6.0 dB bandwidth of the MIMO antenna element is 4.75 GHz to 4.91 GHz,covering 2.2 GHz to 2.4 GHz in the S-band and 4.8 GHz to 4.9 GHz in the 5G band,respectively.In the 5G band,the isolation between any element of the MIMO antenna and the S-band monopole antenna is not less than 21 dB,and the isolation between the MIMO antenna elements is better than 23 dB,indicating t-hat the isolation between the antenna elements is high.The proposed antenna is suitable for the application on UAV airborne platforms.展开更多
Propagation of light beams in turbid media such as underwater environments,fog,clouds,or biological tissues finds increasingly important applications in science and technology,including bio-imaging,underwater communic...Propagation of light beams in turbid media such as underwater environments,fog,clouds,or biological tissues finds increasingly important applications in science and technology,including bio-imaging,underwater communication,and free-space communication technologies.While many of these applications traditionally relied on conventional,linearly polarized Gaussian beams,light possesses many degrees of freedom that are still largely unexplored,such as spin angular momentum(SAM)and orbital angular momentum(OAM).Here,we present nonlinear light-matter interactions of such complex light beams with"rotational"degrees of freedom in engineered nonlinear colloidal media.By making use of both variational and perturbative approach,we consider non-cylindrical optical vortices,elliptical optical vortices,and higher-order Bessel beams integrated in time(HOBBIT)to predict the dynamics and stability of the evolution of these beams.These results may find applications in many scenarios involving light transmission in strongly scattering environments.展开更多
Aiming to reduce the computational costs and converge to global optimum, a novel method is proposed to solve the optimization of a cost function in the estimation of direction of arrival (DOA). In this method, a gen...Aiming to reduce the computational costs and converge to global optimum, a novel method is proposed to solve the optimization of a cost function in the estimation of direction of arrival (DOA). In this method, a genetic algorithm (GA) and fuzzy discrete particle swarm optimization (FDPSO) are applied to optimize the direction of arrival and power parameters of the mode simultaneously. Firstly, the GA algorithm is applied to make the solution fall into the global searching. Secondly, the FDPSO method is utilized to narrow down the search field. In FDPSO, a chaotic factor and a crossover method are added to speed up the convergence. This approach has been demonstrated through some computational simulations. It is shown that the proposed algorithm can estimate both the DOA and the powers accurately. It is more efficient than some present methods, such as the Newton-like algorithm, Akaike information critical (AIC), particle swarm optimization (PSO), and genetic algorithm with particle swarm optimization (GA-PSO).展开更多
Dual argon plasmas ignited by one direct current power source are used to treat an aqueous solution of hydrogen tetrachloroaurate-(Ⅲ)trihydrate(HAuCl4·3H2O)which is contained in an H-type electrochemical c...Dual argon plasmas ignited by one direct current power source are used to treat an aqueous solution of hydrogen tetrachloroaurate-(Ⅲ)trihydrate(HAuCl4·3H2O)which is contained in an H-type electrochemical cell.The solution contained in one cell acts as a cathode,and in the other as an anode.Experiments are carried out to directly visualize the formation process of gold nanoparticles(Au NPs)in separated cells of the H-type electrochemical reactor.The results and analyzes suggest that hydrogen peroxide and hydrated electrons generated from the plasma-liquid interactions play the roles of reductants in the solutions,respectively.Hydrogen peroxide can be generated in the case of the liquid being a cathode or an anode,while most of hydrated electrons are formed in the case of the liquid being an anode.Therefore,the reduction of the AuCl4−ions is mostly attributed to the hydrogen peroxide as the liquid acts as a cathode,while to the hydrogen peroxide and hydrated electrons as the liquid acts as an anode.Moreover,the p H value of the solution can be used to tune the formation processes and final form of the Au NPs due to its mediation of reductants.展开更多
Background:We test a deep learning(DL)supported remote diagnosis approach to detect diabetic retinopathy(DR)and other referable retinal pathologies using ultra-wide-field(UWF)Optomap.Methods:Prospective,non-randomized...Background:We test a deep learning(DL)supported remote diagnosis approach to detect diabetic retinopathy(DR)and other referable retinal pathologies using ultra-wide-field(UWF)Optomap.Methods:Prospective,non-randomized study involving diabetic patients seen at endocrinology clinics.Non-expert imagers were trained to obtain non-dilated images using UWF Primary.Images were graded by two retina specialists and classified as DR or incidental retinal findings.Cohen’s kappa was used to test the agreement between the remote diagnosis and the gold standard exam.A novel DL model was trained to identify the presence or absence of referable pathology,and sensitivity,specificity and area under the receiver operator characteristics curve(AUROC)were used to assess its performance.Results:A total of 265 patients were enrolled,of which 241 patients were imaged(433 eyes).The mean age was 50±17 years,45%of patients were female,34%had a diagnosis of diabetes mellitus type 1,and 66%of type 2.The average Hemoglobin A1c was 8.8±2.3%,and 81%were on Insulin.Of the 433 images,404(93%)were gradable,64 patients(27%)were referred to a retina specialist,and 46(19%)were referred to comprehensive ophthalmologist for a referable retinal pathology on remote diagnosis.Cohen’s kappa was 0.58,indicating moderate agreement.Our DL algorithm achieved an accuracy of 82.8%(95%CI:80.3-85.2%),a sensitivity of 81.0%(95%CI:78.5-83.6%),specificity of 73.5%(95%CI:70.6-76.3%),and AUROC of 81.0%(95%CI:78.5-83.6%).Conclusions:UWF Primary can be used in the non-ophthalmology setting to screen for referable retinal pathology and can be successfully supported by an automated algorithm for image classification.展开更多
Orthogonal matrices have become a vital means for coding and signal processing owing to their unique distributional properties.Although orthogonal matrices based on amplitude or phase combinations have been extensivel...Orthogonal matrices have become a vital means for coding and signal processing owing to their unique distributional properties.Although orthogonal matrices based on amplitude or phase combinations have been extensively explored,the orthogonal matrix of polarization combinations(OMPC)is a novel,relatively unexplored concept.Herein,we propose a method for constructing OMPCs of any dimension encompassing 4n(where n is 1,2,4,8,…)mutually orthogonal 2ncomponent polarization combinations.In the field of holography,the integration of polarization multiplexing techniques with polarization-sensitive materials is expected to emerge as a groundbreaking approach for multichannel hologram multiplexing,offering considerable enhancements in data storage capacity and security.A multidimensional OMPC enables the realization of multichannel multiplexing and dynamical modulation of information in polarization holographic recording.Despite consolidating all information into a single position within the material,we effectively avoided extraneous crosstalk during the reconstruction process.Our results show that achieving four distinct holographic images individually and simultaneously depends on the polarization combination represented by the incident wave.This discovery opens up a new avenue for achieving highly holographic information storage and dynamically displayed information,harnessing the potential of OMPC to expand the heretofore limited dimensionality of orthogonal polarization.展开更多
With the rapid development of mobile wireless Internet and high-precision localization devices,location-based services(LBS)bring more convenience for people over recent years.In LBS,if the original location data are d...With the rapid development of mobile wireless Internet and high-precision localization devices,location-based services(LBS)bring more convenience for people over recent years.In LBS,if the original location data are directly provided,serious privacy problems raise.As a response to these problems,a large number of location-privacy protection mechanisms(LPPMs)(including formal LPPMs,FLPPMs,etc.)and their evaluation metrics have been proposed to prevent personal location information from being leakage and quantify privacy leakage.However,existing schemes independently consider FLPPMs and evaluation metrics,without synergizing them into a unifying framework.In this paper,a unified model is proposed to synergize FLPPMs and evaluation metrics.In detail,the probabilistic process calculus(calledδ-calculus)is proposed to characterize obfuscation schemes(which is a LPPM)and integrateα-entropy toδ-calculus to evaluate its privacy leakage.Further,we use two calculus moving and probabilistic choice to model nodes’mobility and compute its probability distribution of nodes’locations,and a renaming function to model privacy leakage.By formally defining the attacker’s ability and extending relative entropy,an evaluation algorithm is proposed to quantify the leakage of location privacy.Finally,a series of examples are designed to demonstrate the efficiency of our proposed approach.展开更多
There is no reasonable scientific basis for selecting the excellent teachers of the school’s courses.To solve the practical problem,we firstly give a series of normalization models for defining the key attributes of ...There is no reasonable scientific basis for selecting the excellent teachers of the school’s courses.To solve the practical problem,we firstly give a series of normalization models for defining the key attributes of teachers’professional foundation,course difficulty coefficient,and comprehensive evaluation of teaching.Then,we define a partial weight function to calculate the key attributes,and obtain the partial recommendation values.Next,we construct a highly sparse Teaching Recommendation Factorization Machines(TRFMs)model,which takes the 5-tuples relation including teacher,course,teachers’professional foundation,course difficulty,teaching evaluation as the feature vector,and take partial recommendation value as the recommendation label.Finally,we design a novel Top-N excellent teacher recommendation algorithm based on TRFMs by course classification on the highly sparse dataset.Experimental results show that the proposed TRFMs and recommendation algorithm can accurately realize the recommendation of excellent teachers on a highly sparse historical teaching dataset.The recommendation accuracy is superior to that of the three-dimensional tensor decomposition model algorithm which also solves sparse datasets.The proposed method can be used as a new recommendation method applied to the teaching arrangements in all kinds of schools,which can effectively improve the teaching quality.展开更多
We study the interactions of moving discrete solitons in waveguide arrays with two types of point defects that are constructed by varying either the local linear coupling or local waveguide propagation constant at the...We study the interactions of moving discrete solitons in waveguide arrays with two types of point defects that are constructed by varying either the local linear coupling or local waveguide propagation constant at the center of the waveguide array. A broad discrete soliton is kicked toward the defect and interacts with it. Transmission, reflection, scattering, and trapping during the interaction between the soliton and the defect occur depending on the parameters. The detailed behavior of the soliton dynamics is analyzed numerically. A transmission window in the parameter domain is found and the behavior of this window for different parameters is studied. The dynamics of the soliton in the transmission window is found to have chaotic features under certain circumstances and the causes of these phenomena are identified and discussed.展开更多
Multiple resonant excitations of surface plasmons in a graphene stratified slab are realized by Otto configuration at terahertz frequencies. The proposed graphene stratified slab consists of alternating dielectric lay...Multiple resonant excitations of surface plasmons in a graphene stratified slab are realized by Otto configuration at terahertz frequencies. The proposed graphene stratified slab consists of alternating dielectric layers and graphene sheets, and is sandwiched between a prism and another semi-infinite medium. Optical response and field distribution are determined by the transfer matrix method with the surface current density boundary condition.Multiple resonant excitations appear on the angular reflection spectrum, and are analyzed theoretically via the phase-matching condition. Furthermore, the effects of the system parameters are investigated. Among them, the Fermi levels can tune the corresponding resonances independently. The proposed concept can be engineered for promising applications, including angular selective or multiplex filters, multiple channel sensors, and directional delivery of energy.展开更多
The rapid development of optical technologies,including optical trapping,enhanced imaging,and microscopy,necessitates fundamentally new approaches to higher-dimensional optical beam shaping.We introduce a rigorous the...The rapid development of optical technologies,including optical trapping,enhanced imaging,and microscopy,necessitates fundamentally new approaches to higher-dimensional optical beam shaping.We introduce a rigorous theoretical approach for sculpting three-dimensional,topological particle-like objects,such as optical knots or links,including precise control of their individual parts.Universally applicable to knots created using braided zero lines,our method is validated through theoretical analysis and experimental measurements.The proposed approach enables new degrees of freedom in multi-dimensional singularities shaping,including rotations,shifts,and rescaling of their parts for enhanced stability in complex media.These results may find applications in the fields of three-dimensional optical trapping,manipulation,and subwavelength microscopy,as well as probing and imaging through atmospheric or underwater turbulence.展开更多
Dry eye disease(DED)is a widespread ocular surface disorder that requires early and accurate diagnostics for effective management.Among various biomarkers,matrix metalloproteinase-9(MMP-9)has been validated as a relia...Dry eye disease(DED)is a widespread ocular surface disorder that requires early and accurate diagnostics for effective management.Among various biomarkers,matrix metalloproteinase-9(MMP-9)has been validated as a reliable indicator of ocular inflammation.However,current lateral flow assays(LFAs),provide only binary outcomes,which limits their ability to stratify disease severity and guide personalized treatment.More advanced and accessible diagnostic strategies are needed to address this gap.This paper presents a portable smartphone-based platform that integrates machine learning(ML)with LFAs for multiclass detection of tear MMP-9.Customized LFA strips were developed to detect six different concentrations of MMP-9.Images were captured using six smartphones under four distinct backgrounds,generating a dataset of 864 images.The segment anything model(SAM)was applied for strip segmentation,achieving 100%accuracy across all samples.ResNet34 and MobileNet_V2 ML model,pre-trained on ImageNet,were used for classification and achieved accuracies of 99%and over 97%,respectively.The results demonstrate that ML-assisted smartphone imaging enables reliable multiclass detection,overcoming the limitations of conventional binary LFA assays.This approach addresses variability across devices and environments,eliminates the need for external readers,and provides a practical,low-cost solution for DED diagnosis.More broadly,the framework can be extended to other LFA-based assays,supporting scalable point-of-care testing(POCT)in resource-limited settings.展开更多
Recent progress in meta-optics,particularly the development of metasurfaces,has enabled compact,tunable,and integrable control over the amplitude,phase,polarization,and wavefront of light,thereby allowing a paradigm s...Recent progress in meta-optics,particularly the development of metasurfaces,has enabled compact,tunable,and integrable control over the amplitude,phase,polarization,and wavefront of light,thereby allowing a paradigm shift in traditional optical science and engineering.In parallel,the ability to generate and precisely manipulate nonclassical light is central to quantum optics,enabling key functionalities in secure quantum communication,precision measurement,and scalable quantum computation.This review aims to bridge these two rapidly advancing fields by presenting recent progress in the application of structured materials to the manipulation of quantum states of light.展开更多
As cloud computing continues to evolve,managing CPU resources effectively has become a critical task for ensuring system performance and efficiency.Traditional CPU resource management methods,such as static allocation...As cloud computing continues to evolve,managing CPU resources effectively has become a critical task for ensuring system performance and efficiency.Traditional CPU resource management methods,such as static allocation and manual optimization,are increasingly inadequate in handling dynamic,fluctuating workloads characteristic of modern cloud environments.This paper explores the use of Reinforcement Learning(RL)for adaptive CPU resource management,offering a dynamic,data-driven approach to optimizing resource allocation in real-time.Reinforcement learning,particularly Q-learning and Deep Q Networks(DQNs),enables cloud systems to autonomously adjust CPU resources based on workload demands,improving system efficiency and minimizing resource wastage.This paper discusses the key principles of reinforcement learning,its applications in CPU resource management,the benefits of its implementation,and the challenges that need to be addressed for broader adoption.Finally,the paper highlights future directions for integrating RL with other machine learning techniques and its potential impact on cloud infrastructure optimization.展开更多
During the past decades,major advances have been made in both the generation and detection of infrared light;however,its efficient wavefront manipulation and information processing still encounter great challenges.Eff...During the past decades,major advances have been made in both the generation and detection of infrared light;however,its efficient wavefront manipulation and information processing still encounter great challenges.Efficient and fast optoelectronic modulators and spatial light modulators are required for mid-infrared imaging,sensing,security screening,communication and navigation,to name a few.However,their development remains elusive,and prevailing methods reported so far have suffered from drawbacks that significantly limit their practical applications.In this study,by leveraging graphene and metasurfaces,we demonstrate a high-performance free-space mid-infrared modulator operating at gigahertz speeds,low gate voltage and room temperature.We further pixelate the hybrid graphene metasurface to form a prototype spatial light modulator for high frame rate single-pixel imaging,suggesting orders of magnitude improvement over conventional liquid crystal or micromirror-based spatial light modulators.This work opens up the possibility of exploring wavefront engineering for infrared technologies for which fast temporal and spatial modulations are indispensable.展开更多
The addition of surface acoustic wave(SAW)technologies to microfluidics has greatly advanced lab-on-a-chip applications due to their unique and powerful attributes,including high-precision manipulation,versatility,int...The addition of surface acoustic wave(SAW)technologies to microfluidics has greatly advanced lab-on-a-chip applications due to their unique and powerful attributes,including high-precision manipulation,versatility,integrability,biocompatibility,contactless nature,and rapid actuation.However,the development of SAW microfluidic devices is limited by complex and time-consuming microanofabrication techniques and access to cleanroom facilities for multistep photolithography and vacuum-based processing.To simplify the fabrication of SAW microfluidic devices with customizable dimensions and functions,we utilized the additive manufacturing technique of aerosol jet printing.We successfully fabricated customized SAW microfluidic devices of varying materials,including silver nanowires,graphene,and poly(3,4-ethylenedioxythiophene)polystyrene sulfonate(PEDOT:PSS).To characterize and compare the acoustic actuation performance of these aerosol jet printed SAW microfluidic devices with their cleanroom-fabricated counterparts,the wave displacements and resonant frequencies of the different fabricated devices were directly measured through scanning laser Doppler vibrometry.Finally,to exhibit the capability of the aerosol jet printed devices for lab-on-a-chip applications,we successfully conducted acoustic streaming and particle concentration experiments.Overall,we demonstrated a novel solution-based,direct-write,single-step,cleanroom-free additive manufacturing technique to rapidly develop SAW microfluidic devices that shows viability for applications in the fields of biology,chemistry,engineering,and medicine.展开更多
Transformation optics has formulated a versatile framework to mold the flow of light and tailor its spatial characteristics at will.Despite its huge success in bringing scientific fiction(such as invisibility cloaking...Transformation optics has formulated a versatile framework to mold the flow of light and tailor its spatial characteristics at will.Despite its huge success in bringing scientific fiction(such as invisibility cloaking)into reality,the coordinate transformation often yields extreme material parameters unfeasible even with metamaterials.Here,we demonstrate a new transformation paradigm based upon the invariance of the eigenspectra of the Hamiltonian of a physical system,enabled by supersymmetry.By creating a gradient-index metamaterial to control the local index variation in a family of isospectral optical potentials,we demonstrate broadband continuous supersymmetric transformation in optics,on a silicon chip,to simultaneously transform the transverse spatial characteristics of multiple optical states for arbitrary steering and switching of light flows.Through a novel synergy of symmetry physics and metamaterials,our work provides an adaptable strategy to conveniently tame the flow of light with full exploitation of its spatial degree of freedom.展开更多
摘要The rapid advancement of deep learning and the emergence of largescale neural models,such as bidirectional encoder representations from transformers(BERT),generative pre-trained transformer(GPT),and large language model Meta AI(LLaMa),have brought significant computational and energy challenges.Neuromorphic computing presents a biologically inspired approach to addressing these issues,leveraging event-driven processing and in-memory computation for enhanced energy efficiency.This survey explores the intersection of neuromorphic computing and large-scale deep learning models,focusing on neuromorphic models,learning methods,and hardware.We highlight transferable techniques from deep learning to neuromorphic computing and examine the memoryrelated scalability limitations of current neuromorphic systems.Furthermore,we identify potential directions to enable neuromorphic systems to meet the growing demands of modern AI workloads.
基金the National Science Foundations(NSFs)(1822085,1725456,1816833,1500848,1719160,and 1725447)the NSF Computing and Communication Foundations(1740352)+1 种基金the Nanoelectronics COmputing REsearch Program in the Semiconductor Research Corporation(NC-2766-A)the Center for Research in Intelligent Storage and Processing-in-Memory,one of six centers in the Joint University Microelectronics Program,a SRC program sponsored by Defense Advanced Research Projects Agency.
摘要Recently,due to the availability of big data and the rapid growth of computing power,artificial intelligence(AI)has regained tremendous attention and investment.Machine learning(ML)approaches have been successfully applied to solve many problems in academia and in industry.Although the explosion of big data applications is driving the development of ML,it also imposes severe challenges of data processing speed and scalability on conventional computer systems.Computing platforms that are dedicatedly designed for AI applications have been considered,ranging from a complement to von Neumann platforms to a“must-have”and stand-alone technical solution.These platforms,which belong to a larger category named“domain-specific computing,”focus on specific customization for AI.In this article,we focus on summarizing the recent advances in accelerator designs for deep neural networks(DNNs)-that is,DNN accelerators.We discuss various architectures that support DNN executions in terms of computing units,dataflow optimization,targeted network topologies,architectures on emerging technologies,and accelerators for emerging applications.We also provide our visions on the future trend of AI chip designs.
摘要The terahertz(THz) far-field radiation properties of a butterfly-shaped photoconductive antenna(PCA) were experimentally studied using a home-built THz time-domain spectroscopy(THz-TDS) setup.To distinguish the contribution of in-gap photocurrent and antenna structure to far-field radiation,polarization-dependent THz field was measured and quantified as the illuminating laser beam moved along the bias field within the gap region of electrodes. The result suggests that, although the far-field THz radiation originates from the in-gap photocurrent, the antenna structure of butterfly-shaped PCA dominates the overall THz radiation. In addition, to explore the impact of photoconductive material,radiation properties of butterfly-shaped PCAs fabricated on both low-temperature-grown GaAs(LT-GaAs) and semi-insulating GaAs(Si-GaAs) were characterized and compared. Consistent with previous experiments, it is observed that while Si-GaAs-based PCA can emit higher THz field than LT-GaAs-based PCA at low laser power, it would saturate more severely as laser power increased and eventually be surpassed by LT-GaAs-based PCA. Beyond that, it is found the severe saturation effect of Si-GaAs was due to the longer carrier lifetime and higher carrier mobility, which was confirmed by the numerical simulation.
基金supported by the National Natural Science Foundation of China under Grants No.61801406 and No.U23A20651in part by the Research Fund of Sichuan Provincial under Grants No.2024NSFSC0478 and No.2022YFG0259+1 种基金in part by the Research Fund of Key R&D Projects in Sichuan Province under Grant No.2022-ZY00-00009-GXin part by the Mianyang Central Guiding Local Science and Technology Development Fund Project under Grants No.2023ZYDF002 and No.2023ZYDF092.
摘要A dual-band and high-isolation shared-aperture antenna for unmanned aerial vehicle(UAV)platforms has been proposed.This shared-aperture antenna consists of a rectangular monopole antenna and a 4-element multiple input multiple output(MIMO)antenna.In order to increase the isolation,several double split ring metamaterial(MTM)structures are introduced between antenna elements.The antenna radiator and the MTM structure are located on the front and back of the same dielectric substrate,respectively,and are perpendicular to a circular metal plate.The overall size of the antenna substrate is 124 mm×38 mm×1.016 mm.Moreover,the antenna prototype is constructed and measured,and the simulated and measured results are in good agreement.The measured results show that the-10 dB bandwidth of the monopole antenna is 1.92 GHz to 2.75 GHz,and the common-6.0 dB bandwidth of the MIMO antenna element is 4.75 GHz to 4.91 GHz,covering 2.2 GHz to 2.4 GHz in the S-band and 4.8 GHz to 4.9 GHz in the 5G band,respectively.In the 5G band,the isolation between any element of the MIMO antenna and the S-band monopole antenna is not less than 21 dB,and the isolation between the MIMO antenna elements is better than 23 dB,indicating t-hat the isolation between the antenna elements is high.The proposed antenna is suitable for the application on UAV airborne platforms.
基金support from the Office of Naval Research MURI(N00014-20-1-2550)。
摘要Propagation of light beams in turbid media such as underwater environments,fog,clouds,or biological tissues finds increasingly important applications in science and technology,including bio-imaging,underwater communication,and free-space communication technologies.While many of these applications traditionally relied on conventional,linearly polarized Gaussian beams,light possesses many degrees of freedom that are still largely unexplored,such as spin angular momentum(SAM)and orbital angular momentum(OAM).Here,we present nonlinear light-matter interactions of such complex light beams with"rotational"degrees of freedom in engineered nonlinear colloidal media.By making use of both variational and perturbative approach,we consider non-cylindrical optical vortices,elliptical optical vortices,and higher-order Bessel beams integrated in time(HOBBIT)to predict the dynamics and stability of the evolution of these beams.These results may find applications in many scenarios involving light transmission in strongly scattering environments.
摘要Aiming to reduce the computational costs and converge to global optimum, a novel method is proposed to solve the optimization of a cost function in the estimation of direction of arrival (DOA). In this method, a genetic algorithm (GA) and fuzzy discrete particle swarm optimization (FDPSO) are applied to optimize the direction of arrival and power parameters of the mode simultaneously. Firstly, the GA algorithm is applied to make the solution fall into the global searching. Secondly, the FDPSO method is utilized to narrow down the search field. In FDPSO, a chaotic factor and a crossover method are added to speed up the convergence. This approach has been demonstrated through some computational simulations. It is shown that the proposed algorithm can estimate both the DOA and the powers accurately. It is more efficient than some present methods, such as the Newton-like algorithm, Akaike information critical (AIC), particle swarm optimization (PSO), and genetic algorithm with particle swarm optimization (GA-PSO).
基金the Basic Research Program of Science and Technology of Shenzhen, China (No. JCYJ20190809162617137)National Natural Science Foundation of China (No. 52077185) for partial financial support+1 种基金the Australian Research Council (ARC)QUT Center for Materials Science for partial support
摘要Dual argon plasmas ignited by one direct current power source are used to treat an aqueous solution of hydrogen tetrachloroaurate-(Ⅲ)trihydrate(HAuCl4·3H2O)which is contained in an H-type electrochemical cell.The solution contained in one cell acts as a cathode,and in the other as an anode.Experiments are carried out to directly visualize the formation process of gold nanoparticles(Au NPs)in separated cells of the H-type electrochemical reactor.The results and analyzes suggest that hydrogen peroxide and hydrated electrons generated from the plasma-liquid interactions play the roles of reductants in the solutions,respectively.Hydrogen peroxide can be generated in the case of the liquid being a cathode or an anode,while most of hydrated electrons are formed in the case of the liquid being an anode.Therefore,the reduction of the AuCl4−ions is mostly attributed to the hydrogen peroxide as the liquid acts as a cathode,while to the hydrogen peroxide and hydrated electrons as the liquid acts as an anode.Moreover,the p H value of the solution can be used to tune the formation processes and final form of the Au NPs due to its mediation of reductants.
摘要Background:We test a deep learning(DL)supported remote diagnosis approach to detect diabetic retinopathy(DR)and other referable retinal pathologies using ultra-wide-field(UWF)Optomap.Methods:Prospective,non-randomized study involving diabetic patients seen at endocrinology clinics.Non-expert imagers were trained to obtain non-dilated images using UWF Primary.Images were graded by two retina specialists and classified as DR or incidental retinal findings.Cohen’s kappa was used to test the agreement between the remote diagnosis and the gold standard exam.A novel DL model was trained to identify the presence or absence of referable pathology,and sensitivity,specificity and area under the receiver operator characteristics curve(AUROC)were used to assess its performance.Results:A total of 265 patients were enrolled,of which 241 patients were imaged(433 eyes).The mean age was 50±17 years,45%of patients were female,34%had a diagnosis of diabetes mellitus type 1,and 66%of type 2.The average Hemoglobin A1c was 8.8±2.3%,and 81%were on Insulin.Of the 433 images,404(93%)were gradable,64 patients(27%)were referred to a retina specialist,and 46(19%)were referred to comprehensive ophthalmologist for a referable retinal pathology on remote diagnosis.Cohen’s kappa was 0.58,indicating moderate agreement.Our DL algorithm achieved an accuracy of 82.8%(95%CI:80.3-85.2%),a sensitivity of 81.0%(95%CI:78.5-83.6%),specificity of 73.5%(95%CI:70.6-76.3%),and AUROC of 81.0%(95%CI:78.5-83.6%).Conclusions:UWF Primary can be used in the non-ophthalmology setting to screen for referable retinal pathology and can be successfully supported by an automated algorithm for image classification.
基金financial supports from National Key Research and Development Program of China(2018YFA0701800)Fujian Province Major Science and Technology Program(2020HZ01012)+1 种基金National Natural Science Foundation of China(NSFC)(U22A2080)China Scholarship Council(202109107007).
摘要Orthogonal matrices have become a vital means for coding and signal processing owing to their unique distributional properties.Although orthogonal matrices based on amplitude or phase combinations have been extensively explored,the orthogonal matrix of polarization combinations(OMPC)is a novel,relatively unexplored concept.Herein,we propose a method for constructing OMPCs of any dimension encompassing 4n(where n is 1,2,4,8,…)mutually orthogonal 2ncomponent polarization combinations.In the field of holography,the integration of polarization multiplexing techniques with polarization-sensitive materials is expected to emerge as a groundbreaking approach for multichannel hologram multiplexing,offering considerable enhancements in data storage capacity and security.A multidimensional OMPC enables the realization of multichannel multiplexing and dynamical modulation of information in polarization holographic recording.Despite consolidating all information into a single position within the material,we effectively avoided extraneous crosstalk during the reconstruction process.Our results show that achieving four distinct holographic images individually and simultaneously depends on the polarization combination represented by the incident wave.This discovery opens up a new avenue for achieving highly holographic information storage and dynamically displayed information,harnessing the potential of OMPC to expand the heretofore limited dimensionality of orthogonal polarization.
基金This research is supported in part by the National Key Research and Development Program of China(Grant No.2017YFB0803001)in part by the Key research and Development Program for Guangdong Province under grant(Grant No.2019B010136001)+1 种基金in part by the National Natural Science Foundation of China(Grant No.61872100)Guangxi Natural Science Foundation(No.2017GXNSFAA198372).
摘要With the rapid development of mobile wireless Internet and high-precision localization devices,location-based services(LBS)bring more convenience for people over recent years.In LBS,if the original location data are directly provided,serious privacy problems raise.As a response to these problems,a large number of location-privacy protection mechanisms(LPPMs)(including formal LPPMs,FLPPMs,etc.)and their evaluation metrics have been proposed to prevent personal location information from being leakage and quantify privacy leakage.However,existing schemes independently consider FLPPMs and evaluation metrics,without synergizing them into a unifying framework.In this paper,a unified model is proposed to synergize FLPPMs and evaluation metrics.In detail,the probabilistic process calculus(calledδ-calculus)is proposed to characterize obfuscation schemes(which is a LPPM)and integrateα-entropy toδ-calculus to evaluate its privacy leakage.Further,we use two calculus moving and probabilistic choice to model nodes’mobility and compute its probability distribution of nodes’locations,and a renaming function to model privacy leakage.By formally defining the attacker’s ability and extending relative entropy,an evaluation algorithm is proposed to quantify the leakage of location privacy.Finally,a series of examples are designed to demonstrate the efficiency of our proposed approach.
基金This work was supported by the Planning Subject for the 13th Five-Year Plan of Hunan Provincial Educational Sciences under Grant XJK17BXX006,author D.Y,http://gffzzd654ab2606d04e1ahxp9on65b6x9x6wck.ffgz.tsg.suse.edu.cn/.
摘要There is no reasonable scientific basis for selecting the excellent teachers of the school’s courses.To solve the practical problem,we firstly give a series of normalization models for defining the key attributes of teachers’professional foundation,course difficulty coefficient,and comprehensive evaluation of teaching.Then,we define a partial weight function to calculate the key attributes,and obtain the partial recommendation values.Next,we construct a highly sparse Teaching Recommendation Factorization Machines(TRFMs)model,which takes the 5-tuples relation including teacher,course,teachers’professional foundation,course difficulty,teaching evaluation as the feature vector,and take partial recommendation value as the recommendation label.Finally,we design a novel Top-N excellent teacher recommendation algorithm based on TRFMs by course classification on the highly sparse dataset.Experimental results show that the proposed TRFMs and recommendation algorithm can accurately realize the recommendation of excellent teachers on a highly sparse historical teaching dataset.The recommendation accuracy is superior to that of the three-dimensional tensor decomposition model algorithm which also solves sparse datasets.The proposed method can be used as a new recommendation method applied to the teaching arrangements in all kinds of schools,which can effectively improve the teaching quality.
基金Acknowledgements This work was supported by the National Natural Science Foundation of China (Grant Nos. 11104083, 11204089, and 61172011).
摘要We study the interactions of moving discrete solitons in waveguide arrays with two types of point defects that are constructed by varying either the local linear coupling or local waveguide propagation constant at the center of the waveguide array. A broad discrete soliton is kicked toward the defect and interacts with it. Transmission, reflection, scattering, and trapping during the interaction between the soliton and the defect occur depending on the parameters. The detailed behavior of the soliton dynamics is analyzed numerically. A transmission window in the parameter domain is found and the behavior of this window for different parameters is studied. The dynamics of the soliton in the transmission window is found to have chaotic features under certain circumstances and the causes of these phenomena are identified and discussed.
基金National Natural Science Foundation of China(NSFC)(11604276,61601393,11501481)Key Scientific Project of Fujian Province in China(2015H0039)
摘要Multiple resonant excitations of surface plasmons in a graphene stratified slab are realized by Otto configuration at terahertz frequencies. The proposed graphene stratified slab consists of alternating dielectric layers and graphene sheets, and is sandwiched between a prism and another semi-infinite medium. Optical response and field distribution are determined by the transfer matrix method with the surface current density boundary condition.Multiple resonant excitations appear on the angular reflection spectrum, and are analyzed theoretically via the phase-matching condition. Furthermore, the effects of the system parameters are investigated. Among them, the Fermi levels can tune the corresponding resonances independently. The proposed concept can be engineered for promising applications, including angular selective or multiplex filters, multiple channel sensors, and directional delivery of energy.
摘要The rapid development of optical technologies,including optical trapping,enhanced imaging,and microscopy,necessitates fundamentally new approaches to higher-dimensional optical beam shaping.We introduce a rigorous theoretical approach for sculpting three-dimensional,topological particle-like objects,such as optical knots or links,including precise control of their individual parts.Universally applicable to knots created using braided zero lines,our method is validated through theoretical analysis and experimental measurements.The proposed approach enables new degrees of freedom in multi-dimensional singularities shaping,including rotations,shifts,and rescaling of their parts for enhanced stability in complex media.These results may find applications in the fields of three-dimensional optical trapping,manipulation,and subwavelength microscopy,as well as probing and imaging through atmospheric or underwater turbulence.
基金support from general program of the National Natural Science Foundation of China(Grant No.:62573363)the programs of the Natural Science Foundation of Jiangsu Higher Education(Grant No.:24KJB460030).
摘要Dry eye disease(DED)is a widespread ocular surface disorder that requires early and accurate diagnostics for effective management.Among various biomarkers,matrix metalloproteinase-9(MMP-9)has been validated as a reliable indicator of ocular inflammation.However,current lateral flow assays(LFAs),provide only binary outcomes,which limits their ability to stratify disease severity and guide personalized treatment.More advanced and accessible diagnostic strategies are needed to address this gap.This paper presents a portable smartphone-based platform that integrates machine learning(ML)with LFAs for multiclass detection of tear MMP-9.Customized LFA strips were developed to detect six different concentrations of MMP-9.Images were captured using six smartphones under four distinct backgrounds,generating a dataset of 864 images.The segment anything model(SAM)was applied for strip segmentation,achieving 100%accuracy across all samples.ResNet34 and MobileNet_V2 ML model,pre-trained on ImageNet,were used for classification and achieved accuracies of 99%and over 97%,respectively.The results demonstrate that ML-assisted smartphone imaging enables reliable multiclass detection,overcoming the limitations of conventional binary LFA assays.This approach addresses variability across devices and environments,eliminates the need for external readers,and provides a practical,low-cost solution for DED diagnosis.More broadly,the framework can be extended to other LFA-based assays,supporting scalable point-of-care testing(POCT)in resource-limited settings.
基金Army Research Office(W911NF2310057)National Science Foundation(2240562)North Atlantic Treaty Organization(G5984)。
摘要Recent progress in meta-optics,particularly the development of metasurfaces,has enabled compact,tunable,and integrable control over the amplitude,phase,polarization,and wavefront of light,thereby allowing a paradigm shift in traditional optical science and engineering.In parallel,the ability to generate and precisely manipulate nonclassical light is central to quantum optics,enabling key functionalities in secure quantum communication,precision measurement,and scalable quantum computation.This review aims to bridge these two rapidly advancing fields by presenting recent progress in the application of structured materials to the manipulation of quantum states of light.
摘要As cloud computing continues to evolve,managing CPU resources effectively has become a critical task for ensuring system performance and efficiency.Traditional CPU resource management methods,such as static allocation and manual optimization,are increasingly inadequate in handling dynamic,fluctuating workloads characteristic of modern cloud environments.This paper explores the use of Reinforcement Learning(RL)for adaptive CPU resource management,offering a dynamic,data-driven approach to optimizing resource allocation in real-time.Reinforcement learning,particularly Q-learning and Deep Q Networks(DQNs),enables cloud systems to autonomously adjust CPU resources based on workload demands,improving system efficiency and minimizing resource wastage.This paper discusses the key principles of reinforcement learning,its applications in CPU resource management,the benefits of its implementation,and the challenges that need to be addressed for broader adoption.Finally,the paper highlights future directions for integrating RL with other machine learning techniques and its potential impact on cloud infrastructure optimization.
基金the Los Alamos National Laboratory LDRD ProgramAFOSR under contract no.FA9550-12-0491the AFOSR YIP program under Contract no.FA9550-16-1-0183.
摘要During the past decades,major advances have been made in both the generation and detection of infrared light;however,its efficient wavefront manipulation and information processing still encounter great challenges.Efficient and fast optoelectronic modulators and spatial light modulators are required for mid-infrared imaging,sensing,security screening,communication and navigation,to name a few.However,their development remains elusive,and prevailing methods reported so far have suffered from drawbacks that significantly limit their practical applications.In this study,by leveraging graphene and metasurfaces,we demonstrate a high-performance free-space mid-infrared modulator operating at gigahertz speeds,low gate voltage and room temperature.We further pixelate the hybrid graphene metasurface to form a prototype spatial light modulator for high frame rate single-pixel imaging,suggesting orders of magnitude improvement over conventional liquid crystal or micromirror-based spatial light modulators.This work opens up the possibility of exploring wavefront engineering for infrared technologies for which fast temporal and spatial modulations are indispensable.
基金support from the National Science Foundation(CMMI-2104526,CMMI-2243771,and CMMI-2104295)the Department of Energy(DE-NE0009187)+1 种基金the National Institutes of Health(R01GM132603,R01GM144417,and R01GM135486)supported by the National Science Foundation Graduate Research Fellowship under Grant No.2139754 for J.R.and B.S.
摘要The addition of surface acoustic wave(SAW)technologies to microfluidics has greatly advanced lab-on-a-chip applications due to their unique and powerful attributes,including high-precision manipulation,versatility,integrability,biocompatibility,contactless nature,and rapid actuation.However,the development of SAW microfluidic devices is limited by complex and time-consuming microanofabrication techniques and access to cleanroom facilities for multistep photolithography and vacuum-based processing.To simplify the fabrication of SAW microfluidic devices with customizable dimensions and functions,we utilized the additive manufacturing technique of aerosol jet printing.We successfully fabricated customized SAW microfluidic devices of varying materials,including silver nanowires,graphene,and poly(3,4-ethylenedioxythiophene)polystyrene sulfonate(PEDOT:PSS).To characterize and compare the acoustic actuation performance of these aerosol jet printed SAW microfluidic devices with their cleanroom-fabricated counterparts,the wave displacements and resonant frequencies of the different fabricated devices were directly measured through scanning laser Doppler vibrometry.Finally,to exhibit the capability of the aerosol jet printed devices for lab-on-a-chip applications,we successfully conducted acoustic streaming and particle concentration experiments.Overall,we demonstrated a novel solution-based,direct-write,single-step,cleanroom-free additive manufacturing technique to rapidly develop SAW microfluidic devices that shows viability for applications in the fields of biology,chemistry,engineering,and medicine.
基金U.S.Army Research Office(ARO)(W911NF-19-1-0249 and W911NF-18-1-0348)National Science Foundation(NSF)(CMMI-2037097).
摘要Transformation optics has formulated a versatile framework to mold the flow of light and tailor its spatial characteristics at will.Despite its huge success in bringing scientific fiction(such as invisibility cloaking)into reality,the coordinate transformation often yields extreme material parameters unfeasible even with metamaterials.Here,we demonstrate a new transformation paradigm based upon the invariance of the eigenspectra of the Hamiltonian of a physical system,enabled by supersymmetry.By creating a gradient-index metamaterial to control the local index variation in a family of isospectral optical potentials,we demonstrate broadband continuous supersymmetric transformation in optics,on a silicon chip,to simultaneously transform the transverse spatial characteristics of multiple optical states for arbitrary steering and switching of light flows.Through a novel synergy of symmetry physics and metamaterials,our work provides an adaptable strategy to conveniently tame the flow of light with full exploitation of its spatial degree of freedom.