Parkinson’s disease(PD)is a debilitating neurological disorder affecting over 10 million people worldwide.PD classification models using voice signals as input are common in the literature.It is believed that using d...Parkinson’s disease(PD)is a debilitating neurological disorder affecting over 10 million people worldwide.PD classification models using voice signals as input are common in the literature.It is believed that using deep learning algorithms further enhances performance;nevertheless,it is challenging due to the nature of small-scale and imbalanced PD datasets.This paper proposed a convolutional neural network-based deep support vector machine(CNN-DSVM)to automate the feature extraction process using CNN and extend the conventional SVM to a DSVM for better classification performance in small-scale PD datasets.A customized kernel function reduces the impact of biased classification towards the majority class(healthy candidates in our consideration).An improved generative adversarial network(IGAN)was designed to generate additional training data to enhance the model’s performance.For performance evaluation,the proposed algorithm achieves a sensitivity of 97.6%and a specificity of 97.3%.The performance comparison is evaluated from five perspectives,including comparisons with different data generation algorithms,feature extraction techniques,kernel functions,and existing works.Results reveal the effectiveness of the IGAN algorithm,which improves the sensitivity and specificity by 4.05%–4.72%and 4.96%–5.86%,respectively;and the effectiveness of the CNN-DSVM algorithm,which improves the sensitivity by 1.24%–57.4%and specificity by 1.04%–163%and reduces biased detection towards the majority class.The ablation experiments confirm the effectiveness of individual components.Two future research directions have also been suggested.展开更多
Automated grading of dandruff severity is a clinically significant but challenging task due to the inherent ordinal nature of severity levels and the high prevalence of label noise from subjective expert annotations.S...Automated grading of dandruff severity is a clinically significant but challenging task due to the inherent ordinal nature of severity levels and the high prevalence of label noise from subjective expert annotations.Standard classification methods fail to address these dual challenges,limiting their real-world performance.In this paper,a novel,three-phase training framework is proposed that learns a robust ordinal classifier directly from noisy labels.The approach synergistically combines a rank-based ordinal regression backbone with a cooperative,semi-supervised learning strategy to dynamically partition the data into clean and noisy subsets.A hybrid training objective is then employed,applying a supervised ordinal loss to the clean set.The noisy set is simultaneously trained using a dualobjective that combines a semi-supervised ordinal loss with a parallel,label-agnostic contrastive loss.This design allows themodel to learn fromthe entire noisy subset while using contrastive learning to mitigate the risk of error propagation frompotentially corrupt supervision.Extensive experiments on a new,large-scale,multi-site clinical dataset validate our approach.Themethod achieves state-of-the-art performance with 80.71%accuracy and a 76.86%F1-score,significantly outperforming existing approaches,including a 2.26%improvement over the strongest baseline method.This work provides not only a robust solution for a practical medical imaging problem but also a generalizable framework for other tasks plagued by noisy ordinal labels.展开更多
In the rapidly evolving landscape of today’s digital economy,Financial Technology(Fintech)emerges as a trans-formative force,propelled by the dynamic synergy between Artificial Intelligence(AI)and Algorithmic Trading...In the rapidly evolving landscape of today’s digital economy,Financial Technology(Fintech)emerges as a trans-formative force,propelled by the dynamic synergy between Artificial Intelligence(AI)and Algorithmic Trading.Our in-depth investigation delves into the intricacies of merging Multi-Agent Reinforcement Learning(MARL)and Explainable AI(XAI)within Fintech,aiming to refine Algorithmic Trading strategies.Through meticulous examination,we uncover the nuanced interactions of AI-driven agents as they collaborate and compete within the financial realm,employing sophisticated deep learning techniques to enhance the clarity and adaptability of trading decisions.These AI-infused Fintech platforms harness collective intelligence to unearth trends,mitigate risks,and provide tailored financial guidance,fostering benefits for individuals and enterprises navigating the digital landscape.Our research holds the potential to revolutionize finance,opening doors to fresh avenues for investment and asset management in the digital age.Additionally,our statistical evaluation yields encouraging results,with metrics such as Accuracy=0.85,Precision=0.88,and F1 Score=0.86,reaffirming the efficacy of our approach within Fintech and emphasizing its reliability and innovative prowess.展开更多
This study presents a systematic review of applications of artificial intelligence(abbreviated as AI)and blockchain in supply chain provenance traceability and legal forensics cover five sectors:integrated circuits(ab...This study presents a systematic review of applications of artificial intelligence(abbreviated as AI)and blockchain in supply chain provenance traceability and legal forensics cover five sectors:integrated circuits(abbreviated as ICs),pharmaceuticals,electric vehicles(abbreviated as EVs),drones(abbreviated as UAVs),and robotics—in response to rising trade tensions and geopolitical conflicts,which have heightened concerns over product origin fraud and information security.While previous literature often focuses on single-industry contexts or isolated technologies,this reviewcomprehensively surveys these sectors and categorizes 116 peer-reviewed studies by application domain,technical architecture,and functional objective.Special attention is given to traceability control mechanisms,data integrity,and the use of forensic technologies to detect origin fraud.The study further evaluates real-world implementations,including blockchain-enabled drug tracking systems,EV battery raw material traceability,and UAV authentication frameworks,demonstrating the practical value of these technologies.By identifying technological challenges and policy implications,this research provides a comprehensive foundation for future academic inquiry,industrial adoption,and regulatory development aimed at enhancing transparency,resilience,and trust in global supply chains.展开更多
In recent years,cyber threats have escalated across diverse sectors,with cybercrime syndicates increasingly exploiting system vulnerabilities.Traditional passive defense mechanisms have proven insufficient,particularl...In recent years,cyber threats have escalated across diverse sectors,with cybercrime syndicates increasingly exploiting system vulnerabilities.Traditional passive defense mechanisms have proven insufficient,particularly as Linux platforms—historically overlooked in favor of Windows—have emerged as frequent targets.According to Trend Micro,there has been a substantial increase in Linux-targeted malware,with ransomware attacks on Linux surpassing those on macOS.This alarming trend underscores the need for detection strategies specifically designed for Linux environments.To address this challenge,this study proposes a comprehensive malware detection framework tailored for Linux systems,integrating dynamic behavioral analysis with the semantic reasoning capabilities of large language models(LLMs).Malware samples are executed within sandbox environments to extract behavioral features such as system calls and command-line executions.These features are then systematically mapped to the MITRE ATT&CK framework,incorporating its defined data sources,data components,and Tactics,Techniques,and Procedures(TTPs).Two mapping constructs—Conceptual Definition Mapping and TTP Technical Keyword Mapping—are developed from official MITRE documentation.These resources are utilized to fine-tune an LLM,enabling it to semantically interpret complex behavioral patterns and infer associated attack techniques,including those employed by previously unknown malware variants.The resulting detection pipeline effectively bridges raw behavioral data with structured threat intelligence.Experimental evaluations confirm the efficacy of the proposed system,with the fine-tuned Gemma 2B model demonstrating significantly enhanced accuracy in associating behavioral features with ATT&CK-defined techniques.This study contributes a fully integrated Linux-specific detection framework,a novel approach for transforming unstructured behavioral data into actionable intelligence,improved interpretability of malicious behavior,and a scalable training process for future applications of LLMs in cybersecurity.展开更多
As is known,centralized federated learning faces risks of a single point of failure and privacy breaches,and blockchain-based federated learning frameworks can address these challenges to a certain extent in recent wo...As is known,centralized federated learning faces risks of a single point of failure and privacy breaches,and blockchain-based federated learning frameworks can address these challenges to a certain extent in recent works.However,malicious clients may still illegally access the blockchain to upload malicious data or steal on-chain data.In addition,blockchain-based federated training suffers from a heavy storage burden and excessive network communication overhead.To address these issues,we propose an asynchronous,tiered federated learning storage scheme based on blockchain and IPFS.It manages the execution of federated learning tasks through smart contracts deployed on the blockchain,decentralizing the entire training process.Additionally,the scheme employs a secure and efficient blockchain-based asynchronous tiered architecture,integrating attribute-based access control technology for resource exchange between the clients and the blockchain network.It dynamically manages access control policies during training and adopts a hybrid data storage strategy combining blockchain and IPFS.Experiments with multiple sets of image classification tasks are conducted,indicating that the storage strategy used in this scheme saves nearly 50 percent of the communication overhead and significantly reduces the on-chain storage burden compared to the traditional blockchain-only storage strategy.In terms of training effectiveness,it maintains similar accuracy as centralized training and minimizes the probability of being attacked.展开更多
To transmit customer power data collected by smart meters(SMs)to utility companies,data must first be transmitted to the corresponding data aggregation point(DAP)of the SM.The number of DAPs installed and the installa...To transmit customer power data collected by smart meters(SMs)to utility companies,data must first be transmitted to the corresponding data aggregation point(DAP)of the SM.The number of DAPs installed and the installation location greatly impact the whole network.For the traditional DAP placement algorithm,the number of DAPs must be set in advance,but determining the best number of DAPs is difficult,which undoubtedly reduces the overall performance of the network.Moreover,the excessive gap between the loads of different DAPs is also an important factor affecting the quality of the network.To address the above problems,this paper proposes a DAP placement algorithm,APSSA,based on the improved affinity propagation(AP)algorithm and sparrow search(SSA)algorithm,which can select the appropriate number of DAPs to be installed and the corresponding installation locations according to the number of SMs and their distribution locations in different environments.The algorithm adds an allocation mechanism to optimize the subnetwork in the SSA.APSSA is evaluated under three different areas and compared with other DAP placement algorithms.The experimental results validated that the method in this paper can reduce the network cost,shorten the average transmission distance,and reduce the load gap.展开更多
As smart manufacturing and Industry 4.0 continue to evolve,fault diagnosis of mechanical equipment has become crucial for ensuring production safety and optimizing equipment utilization.To address the challenge of cro...As smart manufacturing and Industry 4.0 continue to evolve,fault diagnosis of mechanical equipment has become crucial for ensuring production safety and optimizing equipment utilization.To address the challenge of cross-domain adaptation in intelligent diagnostic models under varying operational conditions,this paper introduces the CNN-1D-KAN model,which combines a 1D Convolutional Neural Network(1D-CNN)with a Kolmogorov–Arnold Network(KAN).The novelty of this approach lies in replacing the traditional 1D-CNN’s final fully connected layer with a KANLinear layer,leveraging KAN’s advanced nonlinear processing and function approximation capabilities while maintaining the simplicity of linear transformations.Experimental results on the CWRU dataset demonstrate that,under stable load conditions,the CNN-1D-KAN model achieves high accuracy,averaging 96.67%.Furthermore,the model exhibits strong transfer generalization and robustness across varying load conditions,sustaining an average accuracy of 90.21%.When compared to traditional neural networks(e.g.,1D-CNN and Multi-Layer Perceptron)and other domain adaptation models(e.g.,KAN Convolutions and KAN),the CNN-1D-KAN consistently outperforms in both accuracy and F1 scores across diverse load scenarios.Particularly in handling complex cross-domain data,it excels in diagnostic performance.This study provides an effective solution for cross-domain fault diagnosis in Industrial Internet systems,offering a theoretical foundation to enhance the reliability and stability of intelligent manufacturing processes,thus supporting the future advancement of industrial IoT applications.展开更多
Digital watermarking must balance imperceptibility,robustness,complexity,and security.To address the challenge of computational efficiency in trellis-based informed embedding,we propose a modified watermarking framewo...Digital watermarking must balance imperceptibility,robustness,complexity,and security.To address the challenge of computational efficiency in trellis-based informed embedding,we propose a modified watermarking framework that integrates fuzzy c-means(FCM)clustering into the generation off block codewords for labeling trellis arcs.The system incorporates a parallel trellis structure,controllable embedding parameters,and a novel informed embedding algorithm with reduced complexity.Two types of embedding schemes—memoryless and memory-based—are designed to flexibly trade-off between imperceptibility and robustness.Experimental results demonstrate that the proposed method outperforms existing approaches in bit error rate(BER)and computational complexity under various attacks,including additive noise,filtering,JPEG compression,cropping,and rotation.The integration of FCM enhances robustness by increasing the codeword distance,while preserving perceptual quality.Overall,the proposed framework is suitable for real-time and secure watermarking applications.展开更多
Background:Interpersonal alienation has increasingly been recognized as a salient risk factor affecting university students’psychological adjustment and life satisfaction.Guided by Social Identity and Self-Categoriza...Background:Interpersonal alienation has increasingly been recognized as a salient risk factor affecting university students’psychological adjustment and life satisfaction.Guided by Social Identity and Self-Categorization theories,this study examines how alienation influences life satisfaction through the mediating roles of social identity and hope.Methods:This study surveyed 492 Taiwan resident,China undergraduate students(53.7 percent female,mean age 21.08 years)from 60 universities using convenience sampling in May 2023.Data were collected through an online questionnaire distributed via faculty-managed teaching media platforms.Measures included perceived social identity,state hope,interpersonal alienation,and life satisfaction.All instrumentswere adapted from validated scales,translated into traditional Chinese through back-translation,and reviewed by experts to ensure content validity and cultural relevance.Statistical analyses were conducted using SPSS 20 and SmartPLS 4.0.Results:Harman’s single-factor test indicated no significant common method bias.Measurement model analyses demonstrated satisfactory reliability,convergent validity,and absence of multicollinearity.All four hypothesized paths were supported:interpersonal alienation negatively predicted life satisfaction,with perceived social identity and hope serving as individual and sequential mediators.The model explained 10.5%of the variance in social identity,25.3%in hope,and 49.6%in life satisfaction.Group comparisons revealed that male students reported significantly higher hope and life satisfaction than females,and first-year students experienced greater alienation than upper-level peers.Conclusion:This study elucidates how interpersonal alienation undermines life satisfaction among university students and highlights the protective roles of social identity and hope.Findings underscore the importance of fostering psychological resources that promote resilience and well-being.The results offer practical implications for designing educational programs that enhance students’sense of belonging,optimism,and emotional strength.These insights contribute to a deeper theoretical understanding of the mechanisms linking alienation and life satisfaction and inform strategies to support student adaptation and flourishing in higher education.展开更多
The compact torus injector(CTI)for the central fueling of the EAST tokamak has undergone significant upgrades to enhance its injection capability.During the initial phase of the platform testing phase,EAST-CTI demonst...The compact torus injector(CTI)for the central fueling of the EAST tokamak has undergone significant upgrades to enhance its injection capability.During the initial phase of the platform testing phase,EAST-CTI demonstrated relatively low performance,with a maximum velocity of 150 km s^(−1) and a single compact torus(CT)plasma mass of 90μg[Kong D et al 2023 Plasma Sci.Technol.25065601].These parameters were insufficient for conducting central fueling experiments on the EAST tokamak.Consequently,extensive upgrades were carried out to improve the performance of the EAST-CTI system.The compression region was extended from 280 mm to 700 mm to prevent rapid compression and deceleration of the CT plasma,along with an extension of the acceleration region to further increase the plasma acceleration.The power supply system has also been upgraded.These improvements elevated the operating voltage from 8 kV to 15 kV,increased the discharge current from 120 kA to 300 kA and enabled repetitive operation at a maximum rate of 2 Hz.As a result,significant advances in EAST-CTI performance were achieved,with the maximum velocity increasing to 330 km s^(−1) and the CT plasma density reaching 1.5×10^(22) m^(−3),thereby enhancing the system capability for future fueling experiments on EAST.This study offers valuable insights into CTI modification and the improvement of central fueling systems for prospective fusion reactors.展开更多
Detecting cyber attacks in networks connected to the Internet of Things(IoT)is of utmost importance because of the growing vulnerabilities in the smart environment.Conventional models,such as Naive Bayes and support v...Detecting cyber attacks in networks connected to the Internet of Things(IoT)is of utmost importance because of the growing vulnerabilities in the smart environment.Conventional models,such as Naive Bayes and support vector machine(SVM),as well as ensemble methods,such as Gradient Boosting and eXtreme gradient boosting(XGBoost),are often plagued by high computational costs,which makes it challenging for them to perform real-time detection.In this regard,we suggested an attack detection approach that integrates Visual Geometry Group 16(VGG16),Artificial Rabbits Optimizer(ARO),and Random Forest Model to increase detection accuracy and operational efficiency in Internet of Things(IoT)networks.In the suggested model,the extraction of features from malware pictures was accomplished with the help of VGG16.The prediction process is carried out by the random forest model using the extracted features from the VGG16.Additionally,ARO is used to improve the hyper-parameters of the random forest model of the random forest.With an accuracy of 96.36%,the suggested model outperforms the standard models in terms of accuracy,F1-score,precision,and recall.The comparative research highlights our strategy’s success,which improves performance while maintaining a lower computational cost.This method is ideal for real-time applications,but it is effective.展开更多
Cardiovascular diseases(CVDs)remain one of the foremost causes of death globally;hence,the need for several must-have,advanced automated diagnostic solutions towards early detection and intervention.Traditional auscul...Cardiovascular diseases(CVDs)remain one of the foremost causes of death globally;hence,the need for several must-have,advanced automated diagnostic solutions towards early detection and intervention.Traditional auscultation of cardiovascular sounds is heavily reliant on clinical expertise and subject to high variability.To counter this limitation,this study proposes an AI-driven classification system for cardiovascular sounds whereby deep learning techniques are engaged to automate the detection of an abnormal heartbeat.We employ FastAI vision-learner-based convolutional neural networks(CNNs)that include ResNet,DenseNet,VGG,ConvNeXt,SqueezeNet,and AlexNet to classify heart sound recordings.Instead of raw waveform analysis,the proposed approach transforms preprocessed cardiovascular audio signals into spectrograms,which are suited for capturing temporal and frequency-wise patterns.The models are trained on the PASCAL Cardiovascular Challenge dataset while taking into consideration the recording variations,noise levels,and acoustic distortions.To demonstrate generalization,external validation using Google’s Audio set Heartbeat Sound dataset was performed using a dataset rich in cardiovascular sounds.Comparative analysis revealed that DenseNet-201,ConvNext Large,and ResNet-152 could deliver superior performance to the other architectures,achieving an accuracy of 81.50%,a precision of 85.50%,and an F1-score of 84.50%.In the process,we performed statistical significance testing,such as the Wilcoxon signed-rank test,to validate performance improvements over traditional classification methods.Beyond the technical contributions,the research underscores clinical integration,outlining a pathway in which the proposed system can augment conventional electronic stethoscopes and telemedicine platforms in the AI-assisted diagnostic workflows.We also discuss in detail issues of computational efficiency,model interpretability,and ethical considerations,particularly concerning algorithmic bias stemming from imbalanced datasets and the need for real-time processing in clinical settings.The study describes a scalable,automated system combining deep learning,feature extraction using spectrograms,and external validation that can assist healthcare providers in the early and accurate detection of cardiovascular disease.AI-driven solutions can be viable in improving access,reducing delays in diagnosis,and ultimately even the continued global burden of heart disease.展开更多
The integration of the Internet of Things(IoT)into healthcare systems improves patient care,boosts operational efficiency,and contributes to cost-effective healthcare delivery.However,overcoming several associated cha...The integration of the Internet of Things(IoT)into healthcare systems improves patient care,boosts operational efficiency,and contributes to cost-effective healthcare delivery.However,overcoming several associated challenges,such as data security,interoperability,and ethical concerns,is crucial to realizing the full potential of IoT in healthcare.Real-time anomaly detection plays a key role in protecting patient data and maintaining device integrity amidst the additional security risks posed by interconnected systems.In this context,this paper presents a novelmethod for healthcare data privacy analysis.The technique is based on the identification of anomalies in cloud-based Internet of Things(IoT)networks,and it is optimized using explainable artificial intelligence.For anomaly detection,the Radial Boltzmann Gaussian Temporal Fuzzy Network(RBGTFN)is used in the process of doing information privacy analysis for healthcare data.Remora Colony SwarmOptimization is then used to carry out the optimization of the network.The performance of the model in identifying anomalies across a variety of healthcare data is evaluated by an experimental study.This evaluation suggested that themodel measures the accuracy,precision,latency,Quality of Service(QoS),and scalability of themodel.A remarkable 95%precision,93%latency,89%quality of service,98%detection accuracy,and 96%scalability were obtained by the suggested model,as shown by the subsequent findings.展开更多
Trajectory tracking for nonlinear robotic systems remains a fundamental yet challenging problem in control engineering,particularly when both precision and efficiency must be ensured.Conventional control methods are o...Trajectory tracking for nonlinear robotic systems remains a fundamental yet challenging problem in control engineering,particularly when both precision and efficiency must be ensured.Conventional control methods are often effective for stabilization but may not directly optimize long-term performance.To address this limitation,this study develops an integrated framework that combines optimal control principles with reinforcement learning for a single-link robotic manipulator.The proposed scheme adopts an actor–critic structure,where the critic network approximates the value function associated with the Hamilton–Jacobi–Bellman equation,and the actor network generates near-optimal control signals in real time.This dual adaptation enables the controller to refine its policy online without explicit system knowledge.Stability of the closed-loop system is analyzed through Lyapunov theory,ensuring boundedness of the tracking error.Numerical simulations on the single-link manipulator demonstrate that themethod achieves accurate trajectory followingwhile maintaining lowcontrol effort.The results further showthat the actor–critic learning mechanism accelerates convergence of the control policy compared with conventional optimization-based strategies.This work highlights the potential of reinforcement learning integrated with optimal control for robotic manipulators and provides a foundation for future extensions to more complex multi-degree-of-freedom systems.The proposed controller is further validated in a physics-based virtual Gazebo environment,demonstrating stable adaptation and real-time feasibility.展开更多
With the development of technology,diffusion model-based solvers have shown significant promise in solving Combinatorial Optimization(CO)problems,particularly in tackling Non-deterministic Polynomial-time hard(NP-hard...With the development of technology,diffusion model-based solvers have shown significant promise in solving Combinatorial Optimization(CO)problems,particularly in tackling Non-deterministic Polynomial-time hard(NP-hard)problems such as the Traveling Salesman Problem(TSP).However,existing diffusion model-based solvers typically employ a fixed,uniform noise schedule(e.g.,linear or cosine annealing)across all training instances,failing to fully account for the unique characteristics of each problem instance.To address this challenge,we present GraphGuided Diffusion Solvers(GGDS),an enhanced method for improving graph-based diffusion models.GGDS leverages Graph Neural Networks(GNNs)to capture graph structural information embedded in node coordinates and adjacency matrices,dynamically adjusting the noise levels in the diffusion model.This study investigates the TSP by examining two distinct time-step noise generation strategies:cosine annealing and a Neural Network(NN)-based approach.We evaluate their performance across different problem scales,particularly after integrating graph structural information.Experimental results indicate that GGDS outperforms previous methods with average performance improvements of 18.7%,6.3%,and 88.7%on TSP-500,TSP-100,and TSP-50,respectively.Specifically,GGDS demonstrates superior performance on TSP-500 and TSP-50,while its performance on TSP-100 is either comparable to or slightly better than that of previous methods,depending on the chosen noise schedule and decoding strategy.展开更多
Palmprint recognition and palm vein recognition are two emerging biometrics technologies.In the past two decades,many traditional methods have been proposed for palmprint recognition and palm vein recognition and have...Palmprint recognition and palm vein recognition are two emerging biometrics technologies.In the past two decades,many traditional methods have been proposed for palmprint recognition and palm vein recognition and have achieved impressive results.In recent years,in the field of artificial intelligence,deep learning has gradually become the mainstream recognition technology because of its excellent recognition performance.Some researchers have tried to use convolutional neural networks(CNNs)for palmprint recognition and palm vein recognition.However,the architectures of these CNNs have mostly been developed manually by human experts,which is a time-consuming and error-prone process.In order to overcome some shortcomings of manually designed CNN,neural architecture search(NAS)technology has become an important research direction of deep learning.The significance of NAS is to solve the deep learning model's parameter adjustment problem,which is a cross-study combining optimization and machine learning.NAS technology represents the future development direction of deep learning.However,up to now,NAS technology has not been well studied for palmprint recognition and palm vein recognition.In this paper,in order to investigate the problem of NAS-based 2D and 3D palmprint recognition and palm vein recognition in-depth,we conduct a performance evaluation of twenty representative NAS methods on five 2D palmprint databases,two palm vein databases,and one 3D palmprint database.Experimental results show that some NAS methods can achieve promising recognition results.Remarkably,among different evaluated NAS methods,Proxyless NAS achieves the best recognition performance.展开更多
Palmprint recognition and palm vein recognition are two emerging biometrics technologies.In the past two decades,many traditional methods have been proposed for palmprint recognition and palm vein recognition,and have...Palmprint recognition and palm vein recognition are two emerging biometrics technologies.In the past two decades,many traditional methods have been proposed for palmprint recognition and palm vein recognition,and have achieved impressive results.However,the research on deep learningbased palmprint recognition and palm vein recognition is still very preliminary.In this paper,in order to investigate the problem of deep learning based 2D and 3D palmprint recognition and palm vein recognition indepth,we conduct performance evaluation of seventeen representative and classic convolutional neural networks(CNNs)on one 3D palmprint database,five 2D palmprint databases and two palm vein databases.A lot of experiments have been carried out in the conditions of different network structures,different learning rates,and different numbers of network layers.We have also conducted experiments on both separate data mode and mixed data mode.Experimental results show that these classic CNNs can achieve promising recognition results,and the recognition performance of recently proposed CNNs is better.Particularly,among classic CNNs,one of the recently proposed classic CNNs,i.e.,EfficientNet achieves the best recognition accuracy.However,the recognition performance of classic CNNs is still slightly worse than that of some traditional recognition methods.展开更多
Target tracking is one of the most important issues in computer vision and has been applied in many fields of science, engineering and industry. Because of the occlusion during tracking, typical approaches with single...Target tracking is one of the most important issues in computer vision and has been applied in many fields of science, engineering and industry. Because of the occlusion during tracking, typical approaches with single classifier learn much of occluding background information which results in the decrease of tracking performance, and eventually lead to the failure of the tracking algorithm. This paper presents a new correlative classifiers approach to address the above problem. Our idea is to derive a group of correlative classifiers based on sample set method. Then we propose strategy to establish the classifiers and to query the suitable classifiers for the next frame tracking. In order to deal with nonlinear problem, particle filter is adopted and integrated with sample set method. For choosing the target from candidate particles, we define a similarity measurement between particles and sample set. The proposed sample set method includes the following steps. First, we cropped positive samples set around the target and negative samples set far away from the target. Second, we extracted average Haar-like feature from these samples and calculate their statistical characteristic which represents the target model. Third, we define the similarity measurement based on the statistical characteristic of these two sets to judge the similarity between candidate particles and target model. Finally, we choose the largest similarity score particle as the target in the new frame. A number of experiments show the robustness and efficiency of the proposed approach when compared with other state-of-the-art trackers.展开更多
基金The work described in this paper was fully supported by a grant from Hong Kong Metropolitan University(RIF/2021/05).
文摘Parkinson’s disease(PD)is a debilitating neurological disorder affecting over 10 million people worldwide.PD classification models using voice signals as input are common in the literature.It is believed that using deep learning algorithms further enhances performance;nevertheless,it is challenging due to the nature of small-scale and imbalanced PD datasets.This paper proposed a convolutional neural network-based deep support vector machine(CNN-DSVM)to automate the feature extraction process using CNN and extend the conventional SVM to a DSVM for better classification performance in small-scale PD datasets.A customized kernel function reduces the impact of biased classification towards the majority class(healthy candidates in our consideration).An improved generative adversarial network(IGAN)was designed to generate additional training data to enhance the model’s performance.For performance evaluation,the proposed algorithm achieves a sensitivity of 97.6%and a specificity of 97.3%.The performance comparison is evaluated from five perspectives,including comparisons with different data generation algorithms,feature extraction techniques,kernel functions,and existing works.Results reveal the effectiveness of the IGAN algorithm,which improves the sensitivity and specificity by 4.05%–4.72%and 4.96%–5.86%,respectively;and the effectiveness of the CNN-DSVM algorithm,which improves the sensitivity by 1.24%–57.4%and specificity by 1.04%–163%and reduces biased detection towards the majority class.The ablation experiments confirm the effectiveness of individual components.Two future research directions have also been suggested.
文摘Automated grading of dandruff severity is a clinically significant but challenging task due to the inherent ordinal nature of severity levels and the high prevalence of label noise from subjective expert annotations.Standard classification methods fail to address these dual challenges,limiting their real-world performance.In this paper,a novel,three-phase training framework is proposed that learns a robust ordinal classifier directly from noisy labels.The approach synergistically combines a rank-based ordinal regression backbone with a cooperative,semi-supervised learning strategy to dynamically partition the data into clean and noisy subsets.A hybrid training objective is then employed,applying a supervised ordinal loss to the clean set.The noisy set is simultaneously trained using a dualobjective that combines a semi-supervised ordinal loss with a parallel,label-agnostic contrastive loss.This design allows themodel to learn fromthe entire noisy subset while using contrastive learning to mitigate the risk of error propagation frompotentially corrupt supervision.Extensive experiments on a new,large-scale,multi-site clinical dataset validate our approach.Themethod achieves state-of-the-art performance with 80.71%accuracy and a 76.86%F1-score,significantly outperforming existing approaches,including a 2.26%improvement over the strongest baseline method.This work provides not only a robust solution for a practical medical imaging problem but also a generalizable framework for other tasks plagued by noisy ordinal labels.
基金This project was funded by Deanship of Scientific Research(DSR)at King Abdulaziz University,Jeddah underGrant No.(IFPIP-1127-611-1443)the authors,therefore,acknowledge with thanks DSR technical and financial support.
文摘In the rapidly evolving landscape of today’s digital economy,Financial Technology(Fintech)emerges as a trans-formative force,propelled by the dynamic synergy between Artificial Intelligence(AI)and Algorithmic Trading.Our in-depth investigation delves into the intricacies of merging Multi-Agent Reinforcement Learning(MARL)and Explainable AI(XAI)within Fintech,aiming to refine Algorithmic Trading strategies.Through meticulous examination,we uncover the nuanced interactions of AI-driven agents as they collaborate and compete within the financial realm,employing sophisticated deep learning techniques to enhance the clarity and adaptability of trading decisions.These AI-infused Fintech platforms harness collective intelligence to unearth trends,mitigate risks,and provide tailored financial guidance,fostering benefits for individuals and enterprises navigating the digital landscape.Our research holds the potential to revolutionize finance,opening doors to fresh avenues for investment and asset management in the digital age.Additionally,our statistical evaluation yields encouraging results,with metrics such as Accuracy=0.85,Precision=0.88,and F1 Score=0.86,reaffirming the efficacy of our approach within Fintech and emphasizing its reliability and innovative prowess.
文摘This study presents a systematic review of applications of artificial intelligence(abbreviated as AI)and blockchain in supply chain provenance traceability and legal forensics cover five sectors:integrated circuits(abbreviated as ICs),pharmaceuticals,electric vehicles(abbreviated as EVs),drones(abbreviated as UAVs),and robotics—in response to rising trade tensions and geopolitical conflicts,which have heightened concerns over product origin fraud and information security.While previous literature often focuses on single-industry contexts or isolated technologies,this reviewcomprehensively surveys these sectors and categorizes 116 peer-reviewed studies by application domain,technical architecture,and functional objective.Special attention is given to traceability control mechanisms,data integrity,and the use of forensic technologies to detect origin fraud.The study further evaluates real-world implementations,including blockchain-enabled drug tracking systems,EV battery raw material traceability,and UAV authentication frameworks,demonstrating the practical value of these technologies.By identifying technological challenges and policy implications,this research provides a comprehensive foundation for future academic inquiry,industrial adoption,and regulatory development aimed at enhancing transparency,resilience,and trust in global supply chains.
基金supported by the National Science and Technology Council under grant number 113-2221-E-027-126-MY3.
文摘In recent years,cyber threats have escalated across diverse sectors,with cybercrime syndicates increasingly exploiting system vulnerabilities.Traditional passive defense mechanisms have proven insufficient,particularly as Linux platforms—historically overlooked in favor of Windows—have emerged as frequent targets.According to Trend Micro,there has been a substantial increase in Linux-targeted malware,with ransomware attacks on Linux surpassing those on macOS.This alarming trend underscores the need for detection strategies specifically designed for Linux environments.To address this challenge,this study proposes a comprehensive malware detection framework tailored for Linux systems,integrating dynamic behavioral analysis with the semantic reasoning capabilities of large language models(LLMs).Malware samples are executed within sandbox environments to extract behavioral features such as system calls and command-line executions.These features are then systematically mapped to the MITRE ATT&CK framework,incorporating its defined data sources,data components,and Tactics,Techniques,and Procedures(TTPs).Two mapping constructs—Conceptual Definition Mapping and TTP Technical Keyword Mapping—are developed from official MITRE documentation.These resources are utilized to fine-tune an LLM,enabling it to semantically interpret complex behavioral patterns and infer associated attack techniques,including those employed by previously unknown malware variants.The resulting detection pipeline effectively bridges raw behavioral data with structured threat intelligence.Experimental evaluations confirm the efficacy of the proposed system,with the fine-tuned Gemma 2B model demonstrating significantly enhanced accuracy in associating behavioral features with ATT&CK-defined techniques.This study contributes a fully integrated Linux-specific detection framework,a novel approach for transforming unstructured behavioral data into actionable intelligence,improved interpretability of malicious behavior,and a scalable training process for future applications of LLMs in cybersecurity.
基金supported by the National Natural Science Foundation of China(Grant No.52331012)the Natural Science Foundation of Shanghai Municipality(Grant No.21ZR1426500)the Program for Cultivation of Graduate Students’Top-notch Innovative Talents of Shanghai Maritime University(Grant No.2023YBR007).
文摘As is known,centralized federated learning faces risks of a single point of failure and privacy breaches,and blockchain-based federated learning frameworks can address these challenges to a certain extent in recent works.However,malicious clients may still illegally access the blockchain to upload malicious data or steal on-chain data.In addition,blockchain-based federated training suffers from a heavy storage burden and excessive network communication overhead.To address these issues,we propose an asynchronous,tiered federated learning storage scheme based on blockchain and IPFS.It manages the execution of federated learning tasks through smart contracts deployed on the blockchain,decentralizing the entire training process.Additionally,the scheme employs a secure and efficient blockchain-based asynchronous tiered architecture,integrating attribute-based access control technology for resource exchange between the clients and the blockchain network.It dynamically manages access control policies during training and adopts a hybrid data storage strategy combining blockchain and IPFS.Experiments with multiple sets of image classification tasks are conducted,indicating that the storage strategy used in this scheme saves nearly 50 percent of the communication overhead and significantly reduces the on-chain storage burden compared to the traditional blockchain-only storage strategy.In terms of training effectiveness,it maintains similar accuracy as centralized training and minimizes the probability of being attacked.
基金supported by the Fujian University of Technology under Grant GYZ20016,GY-Z18183,and GY-Z19005partially supported by the National Science and Technology Council under Grant NSTC 113-2221-E-224-056-.
文摘To transmit customer power data collected by smart meters(SMs)to utility companies,data must first be transmitted to the corresponding data aggregation point(DAP)of the SM.The number of DAPs installed and the installation location greatly impact the whole network.For the traditional DAP placement algorithm,the number of DAPs must be set in advance,but determining the best number of DAPs is difficult,which undoubtedly reduces the overall performance of the network.Moreover,the excessive gap between the loads of different DAPs is also an important factor affecting the quality of the network.To address the above problems,this paper proposes a DAP placement algorithm,APSSA,based on the improved affinity propagation(AP)algorithm and sparrow search(SSA)algorithm,which can select the appropriate number of DAPs to be installed and the corresponding installation locations according to the number of SMs and their distribution locations in different environments.The algorithm adds an allocation mechanism to optimize the subnetwork in the SSA.APSSA is evaluated under three different areas and compared with other DAP placement algorithms.The experimental results validated that the method in this paper can reduce the network cost,shorten the average transmission distance,and reduce the load gap.
基金supported by the Science and Technology Research Program of Chongqing Municipal Education Commission(Grant Nos.KJQN202100812,KJQN202215901,KJQN202400812).
文摘As smart manufacturing and Industry 4.0 continue to evolve,fault diagnosis of mechanical equipment has become crucial for ensuring production safety and optimizing equipment utilization.To address the challenge of cross-domain adaptation in intelligent diagnostic models under varying operational conditions,this paper introduces the CNN-1D-KAN model,which combines a 1D Convolutional Neural Network(1D-CNN)with a Kolmogorov–Arnold Network(KAN).The novelty of this approach lies in replacing the traditional 1D-CNN’s final fully connected layer with a KANLinear layer,leveraging KAN’s advanced nonlinear processing and function approximation capabilities while maintaining the simplicity of linear transformations.Experimental results on the CWRU dataset demonstrate that,under stable load conditions,the CNN-1D-KAN model achieves high accuracy,averaging 96.67%.Furthermore,the model exhibits strong transfer generalization and robustness across varying load conditions,sustaining an average accuracy of 90.21%.When compared to traditional neural networks(e.g.,1D-CNN and Multi-Layer Perceptron)and other domain adaptation models(e.g.,KAN Convolutions and KAN),the CNN-1D-KAN consistently outperforms in both accuracy and F1 scores across diverse load scenarios.Particularly in handling complex cross-domain data,it excels in diagnostic performance.This study provides an effective solution for cross-domain fault diagnosis in Industrial Internet systems,offering a theoretical foundation to enhance the reliability and stability of intelligent manufacturing processes,thus supporting the future advancement of industrial IoT applications.
基金funded by the National Science and Technology Council,Taiwan,under grant number NSTC 114-2221-E-167-005-MY3,and NSTC 113-2221-E-167-006-.
文摘Digital watermarking must balance imperceptibility,robustness,complexity,and security.To address the challenge of computational efficiency in trellis-based informed embedding,we propose a modified watermarking framework that integrates fuzzy c-means(FCM)clustering into the generation off block codewords for labeling trellis arcs.The system incorporates a parallel trellis structure,controllable embedding parameters,and a novel informed embedding algorithm with reduced complexity.Two types of embedding schemes—memoryless and memory-based—are designed to flexibly trade-off between imperceptibility and robustness.Experimental results demonstrate that the proposed method outperforms existing approaches in bit error rate(BER)and computational complexity under various attacks,including additive noise,filtering,JPEG compression,cropping,and rotation.The integration of FCM enhances robustness by increasing the codeword distance,while preserving perceptual quality.Overall,the proposed framework is suitable for real-time and secure watermarking applications.
基金support of the National Science and Technology Council,Taiwan,grant number NSTC 113-2410-H-018-027,granted to the corresponding author,Dr.Yao-Chung Cheng.
文摘Background:Interpersonal alienation has increasingly been recognized as a salient risk factor affecting university students’psychological adjustment and life satisfaction.Guided by Social Identity and Self-Categorization theories,this study examines how alienation influences life satisfaction through the mediating roles of social identity and hope.Methods:This study surveyed 492 Taiwan resident,China undergraduate students(53.7 percent female,mean age 21.08 years)from 60 universities using convenience sampling in May 2023.Data were collected through an online questionnaire distributed via faculty-managed teaching media platforms.Measures included perceived social identity,state hope,interpersonal alienation,and life satisfaction.All instrumentswere adapted from validated scales,translated into traditional Chinese through back-translation,and reviewed by experts to ensure content validity and cultural relevance.Statistical analyses were conducted using SPSS 20 and SmartPLS 4.0.Results:Harman’s single-factor test indicated no significant common method bias.Measurement model analyses demonstrated satisfactory reliability,convergent validity,and absence of multicollinearity.All four hypothesized paths were supported:interpersonal alienation negatively predicted life satisfaction,with perceived social identity and hope serving as individual and sequential mediators.The model explained 10.5%of the variance in social identity,25.3%in hope,and 49.6%in life satisfaction.Group comparisons revealed that male students reported significantly higher hope and life satisfaction than females,and first-year students experienced greater alienation than upper-level peers.Conclusion:This study elucidates how interpersonal alienation undermines life satisfaction among university students and highlights the protective roles of social identity and hope.Findings underscore the importance of fostering psychological resources that promote resilience and well-being.The results offer practical implications for designing educational programs that enhance students’sense of belonging,optimism,and emotional strength.These insights contribute to a deeper theoretical understanding of the mechanisms linking alienation and life satisfaction and inform strategies to support student adaptation and flourishing in higher education.
基金supported by the National MCF Energy R&D Program of China(Nos.2024YFE03130001 and 2024YFE03130002)the Institute of Energy,Hefei Comprehensive National Science Center(Anhui Energy Laboratory)(Nos.21KZS202 and 23KHH140)+3 种基金the University Synergy Innovation Program of Anhui Province(Nos.GXXT-2021-014 and GXXT-2021-029)National Natural Science Foundation of China(Nos.12105088 and 12305247)the Fundamental Research Funds for the Central Universities of China(No.PA2024GDSK0097)the Anhui Province Key Research and Development Plan Program(Nos.202304a05020006 and 2021006).
文摘The compact torus injector(CTI)for the central fueling of the EAST tokamak has undergone significant upgrades to enhance its injection capability.During the initial phase of the platform testing phase,EAST-CTI demonstrated relatively low performance,with a maximum velocity of 150 km s^(−1) and a single compact torus(CT)plasma mass of 90μg[Kong D et al 2023 Plasma Sci.Technol.25065601].These parameters were insufficient for conducting central fueling experiments on the EAST tokamak.Consequently,extensive upgrades were carried out to improve the performance of the EAST-CTI system.The compression region was extended from 280 mm to 700 mm to prevent rapid compression and deceleration of the CT plasma,along with an extension of the acceleration region to further increase the plasma acceleration.The power supply system has also been upgraded.These improvements elevated the operating voltage from 8 kV to 15 kV,increased the discharge current from 120 kA to 300 kA and enabled repetitive operation at a maximum rate of 2 Hz.As a result,significant advances in EAST-CTI performance were achieved,with the maximum velocity increasing to 330 km s^(−1) and the CT plasma density reaching 1.5×10^(22) m^(−3),thereby enhancing the system capability for future fueling experiments on EAST.This study offers valuable insights into CTI modification and the improvement of central fueling systems for prospective fusion reactors.
基金funded by Institutional Fund Projects under grant no.(IFPDP-261-22)。
文摘Detecting cyber attacks in networks connected to the Internet of Things(IoT)is of utmost importance because of the growing vulnerabilities in the smart environment.Conventional models,such as Naive Bayes and support vector machine(SVM),as well as ensemble methods,such as Gradient Boosting and eXtreme gradient boosting(XGBoost),are often plagued by high computational costs,which makes it challenging for them to perform real-time detection.In this regard,we suggested an attack detection approach that integrates Visual Geometry Group 16(VGG16),Artificial Rabbits Optimizer(ARO),and Random Forest Model to increase detection accuracy and operational efficiency in Internet of Things(IoT)networks.In the suggested model,the extraction of features from malware pictures was accomplished with the help of VGG16.The prediction process is carried out by the random forest model using the extracted features from the VGG16.Additionally,ARO is used to improve the hyper-parameters of the random forest model of the random forest.With an accuracy of 96.36%,the suggested model outperforms the standard models in terms of accuracy,F1-score,precision,and recall.The comparative research highlights our strategy’s success,which improves performance while maintaining a lower computational cost.This method is ideal for real-time applications,but it is effective.
基金funded by the deanship of scientific research(DSR),King Abdulaziz University,Jeddah,under grant No.(G-1436-611-309).
文摘Cardiovascular diseases(CVDs)remain one of the foremost causes of death globally;hence,the need for several must-have,advanced automated diagnostic solutions towards early detection and intervention.Traditional auscultation of cardiovascular sounds is heavily reliant on clinical expertise and subject to high variability.To counter this limitation,this study proposes an AI-driven classification system for cardiovascular sounds whereby deep learning techniques are engaged to automate the detection of an abnormal heartbeat.We employ FastAI vision-learner-based convolutional neural networks(CNNs)that include ResNet,DenseNet,VGG,ConvNeXt,SqueezeNet,and AlexNet to classify heart sound recordings.Instead of raw waveform analysis,the proposed approach transforms preprocessed cardiovascular audio signals into spectrograms,which are suited for capturing temporal and frequency-wise patterns.The models are trained on the PASCAL Cardiovascular Challenge dataset while taking into consideration the recording variations,noise levels,and acoustic distortions.To demonstrate generalization,external validation using Google’s Audio set Heartbeat Sound dataset was performed using a dataset rich in cardiovascular sounds.Comparative analysis revealed that DenseNet-201,ConvNext Large,and ResNet-152 could deliver superior performance to the other architectures,achieving an accuracy of 81.50%,a precision of 85.50%,and an F1-score of 84.50%.In the process,we performed statistical significance testing,such as the Wilcoxon signed-rank test,to validate performance improvements over traditional classification methods.Beyond the technical contributions,the research underscores clinical integration,outlining a pathway in which the proposed system can augment conventional electronic stethoscopes and telemedicine platforms in the AI-assisted diagnostic workflows.We also discuss in detail issues of computational efficiency,model interpretability,and ethical considerations,particularly concerning algorithmic bias stemming from imbalanced datasets and the need for real-time processing in clinical settings.The study describes a scalable,automated system combining deep learning,feature extraction using spectrograms,and external validation that can assist healthcare providers in the early and accurate detection of cardiovascular disease.AI-driven solutions can be viable in improving access,reducing delays in diagnosis,and ultimately even the continued global burden of heart disease.
基金funded by Deanship of Scientific Research(DSR)at King Abdulaziz University,Jeddah under grant No.(RG-6-611-43)the authors,therefore,acknowledge with thanks DSR technical and financial support.
文摘The integration of the Internet of Things(IoT)into healthcare systems improves patient care,boosts operational efficiency,and contributes to cost-effective healthcare delivery.However,overcoming several associated challenges,such as data security,interoperability,and ethical concerns,is crucial to realizing the full potential of IoT in healthcare.Real-time anomaly detection plays a key role in protecting patient data and maintaining device integrity amidst the additional security risks posed by interconnected systems.In this context,this paper presents a novelmethod for healthcare data privacy analysis.The technique is based on the identification of anomalies in cloud-based Internet of Things(IoT)networks,and it is optimized using explainable artificial intelligence.For anomaly detection,the Radial Boltzmann Gaussian Temporal Fuzzy Network(RBGTFN)is used in the process of doing information privacy analysis for healthcare data.Remora Colony SwarmOptimization is then used to carry out the optimization of the network.The performance of the model in identifying anomalies across a variety of healthcare data is evaluated by an experimental study.This evaluation suggested that themodel measures the accuracy,precision,latency,Quality of Service(QoS),and scalability of themodel.A remarkable 95%precision,93%latency,89%quality of service,98%detection accuracy,and 96%scalability were obtained by the suggested model,as shown by the subsequent findings.
基金supported in part by the National Science and Technology Council under Grant NSTC 114-2221-E-027-104.
文摘Trajectory tracking for nonlinear robotic systems remains a fundamental yet challenging problem in control engineering,particularly when both precision and efficiency must be ensured.Conventional control methods are often effective for stabilization but may not directly optimize long-term performance.To address this limitation,this study develops an integrated framework that combines optimal control principles with reinforcement learning for a single-link robotic manipulator.The proposed scheme adopts an actor–critic structure,where the critic network approximates the value function associated with the Hamilton–Jacobi–Bellman equation,and the actor network generates near-optimal control signals in real time.This dual adaptation enables the controller to refine its policy online without explicit system knowledge.Stability of the closed-loop system is analyzed through Lyapunov theory,ensuring boundedness of the tracking error.Numerical simulations on the single-link manipulator demonstrate that themethod achieves accurate trajectory followingwhile maintaining lowcontrol effort.The results further showthat the actor–critic learning mechanism accelerates convergence of the control policy compared with conventional optimization-based strategies.This work highlights the potential of reinforcement learning integrated with optimal control for robotic manipulators and provides a foundation for future extensions to more complex multi-degree-of-freedom systems.The proposed controller is further validated in a physics-based virtual Gazebo environment,demonstrating stable adaptation and real-time feasibility.
基金supported by the National Science and Technology Council,Taiwan,under grant no.NSTC 114-2221-E-197-005-MY3.
文摘With the development of technology,diffusion model-based solvers have shown significant promise in solving Combinatorial Optimization(CO)problems,particularly in tackling Non-deterministic Polynomial-time hard(NP-hard)problems such as the Traveling Salesman Problem(TSP).However,existing diffusion model-based solvers typically employ a fixed,uniform noise schedule(e.g.,linear or cosine annealing)across all training instances,failing to fully account for the unique characteristics of each problem instance.To address this challenge,we present GraphGuided Diffusion Solvers(GGDS),an enhanced method for improving graph-based diffusion models.GGDS leverages Graph Neural Networks(GNNs)to capture graph structural information embedded in node coordinates and adjacency matrices,dynamically adjusting the noise levels in the diffusion model.This study investigates the TSP by examining two distinct time-step noise generation strategies:cosine annealing and a Neural Network(NN)-based approach.We evaluate their performance across different problem scales,particularly after integrating graph structural information.Experimental results indicate that GGDS outperforms previous methods with average performance improvements of 18.7%,6.3%,and 88.7%on TSP-500,TSP-100,and TSP-50,respectively.Specifically,GGDS demonstrates superior performance on TSP-500 and TSP-50,while its performance on TSP-100 is either comparable to or slightly better than that of previous methods,depending on the chosen noise schedule and decoding strategy.
基金supported by National Science Foundation of China(Nos.62076086,61673157,61972129,61972127 and 61702154)Key Research and Development Program in Anhui Province(Nos.202004d07020008 and 201904d07020010)。
文摘Palmprint recognition and palm vein recognition are two emerging biometrics technologies.In the past two decades,many traditional methods have been proposed for palmprint recognition and palm vein recognition and have achieved impressive results.In recent years,in the field of artificial intelligence,deep learning has gradually become the mainstream recognition technology because of its excellent recognition performance.Some researchers have tried to use convolutional neural networks(CNNs)for palmprint recognition and palm vein recognition.However,the architectures of these CNNs have mostly been developed manually by human experts,which is a time-consuming and error-prone process.In order to overcome some shortcomings of manually designed CNN,neural architecture search(NAS)technology has become an important research direction of deep learning.The significance of NAS is to solve the deep learning model's parameter adjustment problem,which is a cross-study combining optimization and machine learning.NAS technology represents the future development direction of deep learning.However,up to now,NAS technology has not been well studied for palmprint recognition and palm vein recognition.In this paper,in order to investigate the problem of NAS-based 2D and 3D palmprint recognition and palm vein recognition in-depth,we conduct a performance evaluation of twenty representative NAS methods on five 2D palmprint databases,two palm vein databases,and one 3D palmprint database.Experimental results show that some NAS methods can achieve promising recognition results.Remarkably,among different evaluated NAS methods,Proxyless NAS achieves the best recognition performance.
基金National Science Foundation of China(Nos.61673157,62076086,61972129 and 61702154)Key Research and Development Program in Anhui Province(Nos.202004d07020008 and 201904d07020010).
文摘Palmprint recognition and palm vein recognition are two emerging biometrics technologies.In the past two decades,many traditional methods have been proposed for palmprint recognition and palm vein recognition,and have achieved impressive results.However,the research on deep learningbased palmprint recognition and palm vein recognition is still very preliminary.In this paper,in order to investigate the problem of deep learning based 2D and 3D palmprint recognition and palm vein recognition indepth,we conduct performance evaluation of seventeen representative and classic convolutional neural networks(CNNs)on one 3D palmprint database,five 2D palmprint databases and two palm vein databases.A lot of experiments have been carried out in the conditions of different network structures,different learning rates,and different numbers of network layers.We have also conducted experiments on both separate data mode and mixed data mode.Experimental results show that these classic CNNs can achieve promising recognition results,and the recognition performance of recently proposed CNNs is better.Particularly,among classic CNNs,one of the recently proposed classic CNNs,i.e.,EfficientNet achieves the best recognition accuracy.However,the recognition performance of classic CNNs is still slightly worse than that of some traditional recognition methods.
基金supported by the National Science Foundation of China(61472289)National Key Research and Development Project(2016YFC0106305)The Key Technology R&D Program of Hubei Provence(2014BAA153)
文摘Target tracking is one of the most important issues in computer vision and has been applied in many fields of science, engineering and industry. Because of the occlusion during tracking, typical approaches with single classifier learn much of occluding background information which results in the decrease of tracking performance, and eventually lead to the failure of the tracking algorithm. This paper presents a new correlative classifiers approach to address the above problem. Our idea is to derive a group of correlative classifiers based on sample set method. Then we propose strategy to establish the classifiers and to query the suitable classifiers for the next frame tracking. In order to deal with nonlinear problem, particle filter is adopted and integrated with sample set method. For choosing the target from candidate particles, we define a similarity measurement between particles and sample set. The proposed sample set method includes the following steps. First, we cropped positive samples set around the target and negative samples set far away from the target. Second, we extracted average Haar-like feature from these samples and calculate their statistical characteristic which represents the target model. Third, we define the similarity measurement based on the statistical characteristic of these two sets to judge the similarity between candidate particles and target model. Finally, we choose the largest similarity score particle as the target in the new frame. A number of experiments show the robustness and efficiency of the proposed approach when compared with other state-of-the-art trackers.
