The rapid growth of IoT networks necessitates efficient Intrusion Detection Systems(IDS)capable of addressing dynamic security threats under constrained resource environments.This paper proposes a hybrid IDS for IoT n...The rapid growth of IoT networks necessitates efficient Intrusion Detection Systems(IDS)capable of addressing dynamic security threats under constrained resource environments.This paper proposes a hybrid IDS for IoT networks,integrating Support Vector Machine(SVM)and Genetic Algorithm(GA)for feature selection and parameter optimization.The GA reduces the feature set from 41 to 7,achieving a 30%reduction in overhead while maintaining an attack detection rate of 98.79%.Evaluated on the NSL-KDD dataset,the system demonstrates an accuracy of 97.36%,a recall of 98.42%,and an F1-score of 96.67%,with a low false positive rate of 1.5%.Additionally,it effectively detects critical User-to-Root(U2R)attacks at a rate of 96.2%and Remote-to-Local(R2L)attacks at 95.8%.Performance tests validate the system’s scalability for networks with up to 2000 nodes,with detection latencies of 120 ms at 65%CPU utilization in small-scale deployments and 250 ms at 85%CPU utilization in large-scale scenarios.Parameter sensitivity analysis enhances model robustness,while false positive examination aids in reducing administrative overhead for practical deployment.This IDS offers an effective,scalable,and resource-efficient solution for real-world IoT system security,outperforming traditional approaches.展开更多
We present a novel method for scale-invariant 3D face recognition by integrating computer-generated holography with the Mellin transform.This approach leverages the scale-invariance property of the Mellin transform to...We present a novel method for scale-invariant 3D face recognition by integrating computer-generated holography with the Mellin transform.This approach leverages the scale-invariance property of the Mellin transform to address challenges related to variations in 3D facial sizes during recognition.By applying the Mellin transform to computer-generated holograms and performing correlation between them,which,to the best of our knowledge,is being done for the first time,we have developed a robust recognition framework capable of managing significant scale variations without compromising recognition accuracy.Digital holograms of 3D faces are generated from a face database,and the Mellin transform is employed to enable robust recognition across scale factors ranging from 0.4 to 2.0.Within this range,the method achieves 100%recognition accuracy,as confirmed by both simulation-based and hybrid optical/digital experimental validations.Numerical calculations demonstrate that our method significantly enhances the accuracy and reliability of 3D face recognition,as evidenced by the sharp correlation peaks and higher peak-to-noise ratio(PNR)values than that of using conventional holograms without the Mellin transform.Additionally,the hybrid optical/digital joint transform correlation hardware further validates the method's effectiveness,demonstrating its capability to accurately identify and distinguish 3D faces at various scales.This work provides a promising solution for advanced biometric systems,especially for those which require 3D scale-invariant recognition.展开更多
Scalable fabrication of homogeneous perovskite films remains crucial for bridging the efficiency gap between lab-scale solar cells and commercial solar modules.To tackle this issue,we introduce NCyano-4-methyl-N-pheny...Scalable fabrication of homogeneous perovskite films remains crucial for bridging the efficiency gap between lab-scale solar cells and commercial solar modules.To tackle this issue,we introduce NCyano-4-methyl-N-phenylbenzenesulfonamide(CMPS) additives into perovskite precursors,enabling slot-die coating of large-area modules under ambient conditions.CMPS suppresses colloidal aggregation and delays crystallization,yielding high-quality uniform films.Small-area devices(0.07 cm^(2) aperture area) incorporating CMPS exhibited a significant efficiency increase from 22.07 % to 24.58 %.Corresponding encapsulated devices maintained 85 % of their initial power conversion efficiency(PCE,average 23.56 %) after 1500 h of continuous maximum power point(MPP) tracking under one-sun illumination at 50-55℃.Furthermore,we demonstrate impressive efficiency of perovskite solar modules,achieving 20.57 %(52 cm^(2) aperture area) for 10 cm × 10 cm mini-modules and 17.02 %(260 cm^(2) aperture area) for 21 cm × 21 cm sub-modules,representing the state-of-the-art performance for solutionprocessed devices at these scales.展开更多
Optimizing routing and resource allocation in decentralized unmanned aerial vehicle(UAV)networks remains challenging due to interference and rapidly changing topologies.The authors introduce a novel framework combinin...Optimizing routing and resource allocation in decentralized unmanned aerial vehicle(UAV)networks remains challenging due to interference and rapidly changing topologies.The authors introduce a novel framework combining double deep Q-networks(DDQNs)and graph neural networks(GNNs)for joint routing and resource allocation.The framework uses GNNs to model the network topology and DDQNs to adaptively control routing and resource allocation,addressing interference and improving network performance.Simulation results show that the proposed approach outperforms traditional methods such as Closest-to-Destination(c2Dst),Max-SINR(mSINR),and Multi-Layer Perceptron(MLP)-based models,achieving approximately 23.5% improvement in throughput,50% increase in connection probability,and 17.6% reduction in number of hops,demonstrating its effectiveness in dynamic UAV networks.展开更多
Federated Learning(FL)has become a leading decentralized solution that enables multiple clients to train a model in a collaborative environment without directly sharing raw data,making it suitable for privacy-sensitiv...Federated Learning(FL)has become a leading decentralized solution that enables multiple clients to train a model in a collaborative environment without directly sharing raw data,making it suitable for privacy-sensitive applications such as healthcare,finance,and smart systems.As the field continues to evolve,the research field has become more complex and scattered,covering different system designs,training methods,and privacy techniques.This survey is organized around the three core challenges:how the data is distributed,how models are synchronized,and how to defend against attacks.It provides a structured and up-to-date review of FL research from 2023 to 2025,offering a unified taxonomy that categorizes works by data distribution(Horizontal FL,Vertical FL,Federated Transfer Learning,and Personalized FL),training synchronization(synchronous and asynchronous FL),optimization strategies,and threat models(data leakage and poisoning attacks).In particular,we summarize the latest contributions in Vertical FL frameworks for secure multi-party learning,communication-efficient Horizontal FL,and domain-adaptive Federated Transfer Learning.Furthermore,we examine synchronization techniques addressing system heterogeneity,including straggler mitigation in synchronous FL and staleness management in asynchronous FL.The survey covers security threats in FL,such as gradient inversion,membership inference,and poisoning attacks,as well as their defense strategies that include privacy-preserving aggregation and anomaly detection.The paper concludes by outlining unresolved issues and highlighting challenges in handling personalized models,scalability,and real-world adoption.展开更多
In contemporary computer vision,convolutional neural networks(CNNs)and vision transformers(ViTs)represent the two primary architectural paradigms for image recognition.While both approaches have been widely adopted in...In contemporary computer vision,convolutional neural networks(CNNs)and vision transformers(ViTs)represent the two primary architectural paradigms for image recognition.While both approaches have been widely adopted in medical imaging applications,they operate based on fundamentally different computational principles.This report attempts to provide brief application notes on ViTs and CNNs,particularly focusing on scenarios that guide the selection of one architecture over the other in practical medical implementations.Generally,CNNs rely on convolutional kernels,localized receptive fields,and weight sharing,enabling efficient hierarchical feature extraction.These properties contribute to strong performance in detecting spatially constrained patterns such as textures,edges,and anatomical boundaries,while maintaining relatively low computational requirements.ViTs,on the other hand,decompose images into smaller segments referred to as tokens and employ self-attention mechanisms to model relationships across the entire image.This global modeling capability allows ViTs to capture long-range dependencies that may be difficult for convolution-based architectures to learn.However,ViTs typically achieve optimal performance when trained on extremely large datasets or when supported by extensive pretraining,as their reduced inductive bias requires greater data exposure to learn robust representations.This report briefly examines the architectural structure,underlying mathematical foundations,and relative performance characteristics of CNNs and ViTs,drawing upon recent findings from contemporary research.Emphasis is placed on understanding how differences in data availability,computational resources,and task requirements influence model effectiveness across medical imaging domains.Most importantly,the report serves as a concise application guide for practitioners seeking informed implementation decisions between these two influential deep learning frameworks.展开更多
This review provides an insightful and comprehensive exploration of the emerging 2D material borophene,both pristine and modified,emphasizing its unique attributes and potential for sustainable applications.Borophene...This review provides an insightful and comprehensive exploration of the emerging 2D material borophene,both pristine and modified,emphasizing its unique attributes and potential for sustainable applications.Borophene’s distinctive properties include its anisotropic crystal structures that contribute to its exceptional mechanical and electronic properties.The material exhibits superior electrical and thermal conductivity,surpassing many other 2D materials.Borophene’s unique atomic spin arrangements further diversify its potential application for magnetism.Surface and interface engineering,through doping,functionalization,and synthesis of hybridized and nanocomposite borophene-based systems,is crucial for tailoring borophene’s properties to specific applications.This review aims to address this knowledge gap through a comprehensive and critical analysis of different synthetic and functionalisation methods,to enhance surface reactivity by increasing active sites through doping and surface modifications.These approaches optimize diffusion pathways improving accessibility for catalytic reactions,and tailor the electronic density to tune the optical and electronic behavior.Key applications explored include energy systems(batteries,supercapacitors,and hydrogen storage),catalysis for hydrogen and oxygen evolution reactions,sensors,and optoelectronics for advanced photonic devices.The key to all these applications relies on strategies to introduce heteroatoms for tuning electronic and catalytic properties,employ chemical modifications to enhance stability and leverage borophene’s conductivity and reactivity for advanced photonics.Finally,the review addresses challenges and proposes solutions such as encapsulation,functionalization,and integration with composites to mitigate oxidation sensitivity and overcome scalability barriers,enabling sustainable,commercial-scale applications.展开更多
Interferometry is a crucial investigative technique used across diverse fields to achieve highprecision measurements.It works by analyzing the phase difference between two interfering waves,which results from variatio...Interferometry is a crucial investigative technique used across diverse fields to achieve highprecision measurements.It works by analyzing the phase difference between two interfering waves,which results from variations in optical path lengths within an interferometer.We introduce a novel method for directly measuring changes in the phase difference within an optical interferometer,importantly,with the added advantage of a controllable enhancement factor.This approach is achieved through a two-step process:first,the optical phase difference is encoded into a sub-GHz radiofrequency(RF)signal using microwave-photonic manipulation;then,RF interferometry-assisted phase amplification is implemented at the destructive interference point.In our experiments,we demonstrate a phase sensitivity of 2.14 rad∕nm operating at 140 MHz using a miniature in-fiber Fabry-Pérot interferometer for sub-nanometer displacement sensing,which reveals a sensitivity magnification factor of 258.6.With further refinement,we anticipate that even higher enhancement factors can be achieved,paving the way for the development of cost-effective,ultrasensitive interferometry-based instruments for high-precision optical measurements.展开更多
In today's connected world,the generation of massive streaming data across diverse domains has become commonplace.In the presence of concept drift,class imbalance,label scarcity,and new class emergence,these chall...In today's connected world,the generation of massive streaming data across diverse domains has become commonplace.In the presence of concept drift,class imbalance,label scarcity,and new class emergence,these challenges jointly degrade representation stability,bias learning toward outdated distributions,and reduce the resilience and reliability of detection in dynamic environments.This paper proposes a streaming classincremental learning(SCIL)framework to address these issues.The SCIL framework integrates an autoencoder(AE)with a multi-layer perceptron for multi-class prediction,employs a dual-loss strategy(classification and reconstruction)for prediction and new class detection,uses corrected pseudo-labels for online training,manages classes with queues,and applies oversampling to handle imbalance.The rationale behind the method's structure is elucidated through ablation studies,and a comprehensive experimental evaluation is performed using both real-world and synthetic datasets that feature class imbalance,incremental classes,and concept drifts.Our results demonstrate that SCIL outperforms strong baselines and state-of-the-art methods.In line with our commitment to Open Science,we make our code and datasets available to the community.展开更多
We discuss recent progress in using machine-learning(ML)-enabled inverse design techniques applied to photonic devices and components.Specifically,we highlight the design of optical sources,including fiber and semicon...We discuss recent progress in using machine-learning(ML)-enabled inverse design techniques applied to photonic devices and components.Specifically,we highlight the design of optical sources,including fiber and semiconductor lasers,as well as Raman and semiconductor optical amplifiers.Although inverse design approaches for optical detectors remain relatively underexplored,we examine optical layers,particularly metamaterial absorbers,as promising candidates for high-performance optical detection.In addition,we underscore advancements in inverse designing passive optical components,including beam splitters,gratings,and optical fibers.These optical blocks are fundamental in developing next-generation standalone optical communication systems and optical sensing networks,including integrated sensing and communication technologies.While categorizing various reported deep learning architectures across five paradigms,we offer a paradigm-based perspective that reveals how different ML techniques function within modern inverse design methods and enable fast,data-driven solutions that significantly reduce design time and computational demands compared with traditional optimization methods.展开更多
In this paper,we propose a fuzzy logic-based coded event-triggered control with self-adjustable prescribed performance(FL-CEC-SPP)to address the trade-off between control performance and communication efficiency in re...In this paper,we propose a fuzzy logic-based coded event-triggered control with self-adjustable prescribed performance(FL-CEC-SPP)to address the trade-off between control performance and communication efficiency in resource-constrained networked control systems.The method integrates a fuzzy-coded event-triggered controller into a coded control framework to dynamically adjust the triggering threshold,thereby reducing unnecessary transmissions while maintaining system stability.A self-adjustable prescribed performance constraint is also incorporated to ensure that the tracking error remains within predefined bounds under arbitrary initial conditions.Theoretical analyses and simulation comparisons show that the method proposed in this paper maintains good tracking performance and stability while reducing the communication burden,and has wide applications in resource-constrained network control systems.展开更多
The hybrid series-parallel microgrid attracts more attention by combining the advantages of both the series-stacked voltage and parallel-expanded capacity.Low-voltage distributed generations(DGs)are connected in serie...The hybrid series-parallel microgrid attracts more attention by combining the advantages of both the series-stacked voltage and parallel-expanded capacity.Low-voltage distributed generations(DGs)are connected in series to form the intra-string,and then multiple strings are interconnected in parallel.For the existing control strategies,both intra-string and inter-string depend on the centralized or distributed control with high communication reliance.It has limited scalability and redundancy under abnormal conditions.Alternatively,in this study,an intra-string distributed and inter-string decentralized control framework is proposed.Within the string,a few DGs close to the AC bus are the leaders to get the string power information and the rest DGs are the followers to acquire the synchronization information through the droop-based distributed consistency.Specifically,the output of the entire string has the active power−angular frequency(ω-P)droop characteristic,and the decentralized control among strings can be autonomously guaranteed.Moreover,the secondary control is designed to realize multi-mode objectives,including on/off-grid mode switching,grid-connected power interactive management,and off-grid voltage quality regulation.As a result,the proposed method has the ability of plug-and-play capabilities,single-point failure redundancy,and seamless mode-switching.Experimental results are provided to verify the effectiveness of the proposed practical solution.展开更多
Reinforcement learning(RL),as an important branch of machine learning,has recently achieved extensive attention and success in many applications.Its main idea is to enable agents to continuously learn to make optimal ...Reinforcement learning(RL),as an important branch of machine learning,has recently achieved extensive attention and success in many applications.Its main idea is to enable agents to continuously learn to make optimal decisions by trying to maximize a reward function for their actions and interactions with the environment.However,making highquality decisions in complex and uncertain real-world scenarios is a challenging task.The interference and attacks in such scenarios tend to destroy the existing strategies.Maintaining RL's optimal performance in various cases and adapting to changing environments remains an important challenge.This article presents a comprehensive review of recent advancements in robust reinforcement learning(RRL),and analyzes them from the perspectives of challenges,methodologies,and applications.It systematically evaluates current progress in RRL and summarizes the commonly used benchmark platforms.Finally,several open challenges are discussed to stimulate further research and guide future developments in this area.展开更多
Accurate blood pressure(BP)monitoring is essential for preventing and managing cardiovascular disease.Advancements in materials science,medicine,flexible electronic,and artificial intelligence(AI)have enabled cuffless...Accurate blood pressure(BP)monitoring is essential for preventing and managing cardiovascular disease.Advancements in materials science,medicine,flexible electronic,and artificial intelligence(AI)have enabled cuffless,unobtrusive BP monitoring systems,offering an alternative to traditional sphygmomanometers.However,extending these advances to real-world cardiovascular care particularly in resource-limited settings remains challenging due to constraints in computational resources,power efficiency,and deployment scalability.This review presents a comprehensive synthesis of AI-enhanced wearable BP monitoring,emphasizing its potential for personalized,scalable,and accessible healthcare.We systematically analyze the end-to-end system architecture,from mechano-electric sensing principles and AI-based estimation models to edge-aware deployment strategies tailored for low-resource environments.We further discuss clinical validation metrics and implementation barriers and prospective strategies.To bridge lab-to-field translation,we propose an innovative"sensor-model-deployment-assessment"co-design framework.This roadmap highlights how AI-enhanced BP technologies can support proactive hypertension control and promote cardiovascular health equity on a global scale.展开更多
Colorectal cancer is the third most diagnosed cancer worldwide,and immune checkpoint inhibitors have shown promising therapeutic outcomes in selected patient groups.This study performed a comprehensive analysis of mul...Colorectal cancer is the third most diagnosed cancer worldwide,and immune checkpoint inhibitors have shown promising therapeutic outcomes in selected patient groups.This study performed a comprehensive analysis of multi-omics data from The Cancer Genome Atlas colorectal adenocarcinoma cohort(TCGA-COADREAD),accessed through cBioPortal,to develop machine learning models for predicting progression-free survival(PFS)following immunotherapy.The dataset included clinical variables,genomic alterations in Kirsten Rat Sarcoma Viral Oncogene Homolog(KRAS),B-Raf Proto-Oncogene(BRAF),and Neuroblastoma RAS Viral Oncogene Homolog(NRAS),microsatellite instability(MSI)status,tumor mutation burden(TMB),and expression of immune checkpoint genes.Kaplan–Meier analysis showed that KRAS mutations were significantly associated with reduced PFS,while BRAF and NRAS mutations had no significant impact.MSI-high tumors exhibited elevated TMB and increased immune checkpoint expression,reflecting their immunologically active phenotype.We developed both survival and classification models,with the Extra Trees classifier achieving the best performance(accuracy=0.86,precision=0.67,recall=0.70,F1-score=0.68,AUC=0.84).These findings highlight the potential of combining genomic and immune biomarkers with machine learning to improve patient stratification and guide personalized immunotherapy decisions.An interactive web application was also developed to enable clinicians to input patient-specific molecular and clinical data and visualize individualized PFS predictions,supporting timely,data-driven treatment planning.展开更多
Textile electronics with extraordinary sensing capabilities holds significant potential in the Artificial Intelligence of Things(AIoT).However,little effort is paid to their mutual advantages of robust interfacial int...Textile electronics with extraordinary sensing capabilities holds significant potential in the Artificial Intelligence of Things(AIoT).However,little effort is paid to their mutual advantages of robust interfacial interactions,ultra-strong mechanical performance,and stability.Herein,we fabricate homogeneous and multifunctional core-shell macrofibers by integrating bridge-functionalized MXene/PEDOT:PSS conductive ink with aligned bacterial cellulose(BC).These resulting macrofibers feature mechanical properties(tensile strength of 433.2 MPa and the Young's modulus of 25.9 GPa),exceptional electrical conductivity(10.05 S cm^(-1))and durable hydrophobicity.Such superior robustness allows for the fabrication of the macrofibers woven into textile-based triboelectric nanogenerator(PKT-TENG)and shows an impressive high-performance of a maximum open-circuit voltage of 272.54 V,short-circuit current of 14.56μA and power density of 86.29 mW m^(-2),which successfully powers commercial electronics.As the proof-of-concept illustration,the macrofibers with durable hydrophobicity and high piezoresistive sensitivity are further employed for precepting diverse liquids that can simultaneously monitor their distinctive motion features via real-time resistance variation on the textile-based array.This work is expected to offer new insights into the design of advanced fibers with ultra-strong mechanical capabilities and high conductivity and provide an avenue for the development of textile electronics for high-performance sensing and intelligent manufacturing.展开更多
Despite the intrinsic durability of polymeric hole transport materials,poly-triarylamines(PTAA)-based inverted perovskite solar cells(PSCs)have lagged behind their counterparts in efficiency,primarily due to poor surf...Despite the intrinsic durability of polymeric hole transport materials,poly-triarylamines(PTAA)-based inverted perovskite solar cells(PSCs)have lagged behind their counterparts in efficiency,primarily due to poor surface wettability,insufficient interfacial contact,and unfavorable energy level alignment at the PTAA/perovskite interface.Here,we report a highly effective interfacial engineering strategy employing the ionic liquid 1,3-dimethylimidazolium dimethyl phosphate(DMIMPH)as a multifunctional interfacial modifier.The incorporation of DMIMPH improves PTAA wettability,promoting the growth of high-quality perovskite films with enhanced interfacial contact.Concurrently,DMIMPH effectively tunes the energy levels of PTAA,enhances its electrical conductivity,and passivates interfacial defects with more efficient hole extraction and charge transport.Moreover,its interaction with residual PbI_(2) modulates perovskite crystallization kinetics,yielding highly crystalline perovskite films with enlarged grain sizes,reduced PbI_(2) residue,and suppressed trap densities.As a result,PTAA-based p-i-n PSCs employing this approach achieve a record certified power conversion efficiency(PCE)of 24.52%,with a champion efficiency of 25.12%—the highest certified value for PTAA-based perovskite devices to date.Impressively,the DMIMPH-modified PSCs without encapsulation maintained 87.48%of their initial efficiency after 1600 h in air.This strategy offers an effective pathway for advancing the performance and stability of polymer-based inverted PSCs.展开更多
Over the past years,many efforts have been accomplished to achieve fast and accurate meta-heuristic algorithms to optimize a variety of real-world problems.This study presents a new optimization method based on an unu...Over the past years,many efforts have been accomplished to achieve fast and accurate meta-heuristic algorithms to optimize a variety of real-world problems.This study presents a new optimization method based on an unusual geological phenomenon in nature,named Geyser inspired Algorithm(GEA).The mathematical modeling of this geological phenomenon is carried out to have a better understanding of the optimization process.The efficiency and accuracy of GEA are verified using statistical examination and convergence rate comparison on numerous CEC 2005,CEC 2014,CEC 2017,and real-parameter benchmark functions.Moreover,GEA has been applied to several real-parameter engineering optimization problems to evaluate its effectiveness.In addition,to demonstrate the applicability and robustness of GEA,a comprehensive investigation is performed for a fair comparison with other standard optimization methods.The results demonstrate that GEA is noticeably prosperous in reaching the optimal solutions with a high convergence rate in comparison with other well-known nature-inspired algorithms,including ABC,BBO,PSO,and RCGA.Note that the source code of the GEA is publicly available at https://www.optim-app.com/projects/gea.展开更多
Electrochemical carbon dioxide reduction reaction(CO_(2)RR)provides a promising way to convert CO_(2)to chemicals.The multicarbon(C_(2+))products,especially ethylene,are of great interest due to their versatile indust...Electrochemical carbon dioxide reduction reaction(CO_(2)RR)provides a promising way to convert CO_(2)to chemicals.The multicarbon(C_(2+))products,especially ethylene,are of great interest due to their versatile industrial applications.However,selectively reducing CO_(2)to ethylene is still challenging as the additional energy required for the C–C coupling step results in large overpotential and many competing products.Nonetheless,mechanistic understanding of the key steps and preferred reaction pathways/conditions,as well as rational design of novel catalysts for ethylene production have been regarded as promising approaches to achieving the highly efficient and selective CO_(2)RR.In this review,we first illustrate the key steps for CO_(2)RR to ethylene(e.g.,CO_(2)adsorption/activation,formation of~*CO intermediate,C–C coupling step),offering mechanistic understanding of CO_(2)RR conversion to ethylene.Then the alternative reaction pathways and conditions for the formation of ethylene and competitive products(C_1 and other C_(2+)products)are investigated,guiding the further design and development of preferred conditions for ethylene generation.Engineering strategies of Cu-based catalysts for CO_(2)RR-ethylene are further summarized,and the correlations of reaction mechanism/pathways,engineering strategies and selectivity are elaborated.Finally,major challenges and perspectives in the research area of CO_(2)RR are proposed for future development and practical applications.展开更多
DURING our discussion at workshops for writing“What Does ChatGPT Say:The DAO from Algorithmic Intelligence to Linguistic Intelligence”[1],we had expected the next milestone for Artificial Intelligence(AI)would be in...DURING our discussion at workshops for writing“What Does ChatGPT Say:The DAO from Algorithmic Intelligence to Linguistic Intelligence”[1],we had expected the next milestone for Artificial Intelligence(AI)would be in the direction of Imaginative Intelligence(II),i.e.,something similar to automatic wordsto-videos generation or intelligent digital movies/theater technology that could be used for conducting new“Artificiofactual Experiments”[2]to replace conventional“Counterfactual Experiments”in scientific research and technical development for both natural and social studies[2]-[6].Now we have OpenAI’s Sora,so soon,but this is not the final,actually far away,and it is just the beginning.展开更多
文摘The rapid growth of IoT networks necessitates efficient Intrusion Detection Systems(IDS)capable of addressing dynamic security threats under constrained resource environments.This paper proposes a hybrid IDS for IoT networks,integrating Support Vector Machine(SVM)and Genetic Algorithm(GA)for feature selection and parameter optimization.The GA reduces the feature set from 41 to 7,achieving a 30%reduction in overhead while maintaining an attack detection rate of 98.79%.Evaluated on the NSL-KDD dataset,the system demonstrates an accuracy of 97.36%,a recall of 98.42%,and an F1-score of 96.67%,with a low false positive rate of 1.5%.Additionally,it effectively detects critical User-to-Root(U2R)attacks at a rate of 96.2%and Remote-to-Local(R2L)attacks at 95.8%.Performance tests validate the system’s scalability for networks with up to 2000 nodes,with detection latencies of 120 ms at 65%CPU utilization in small-scale deployments and 250 ms at 85%CPU utilization in large-scale scenarios.Parameter sensitivity analysis enhances model robustness,while false positive examination aids in reducing administrative overhead for practical deployment.This IDS offers an effective,scalable,and resource-efficient solution for real-world IoT system security,outperforming traditional approaches.
基金financial supports from the National Natural Science Foundation of China(Grant No.6227511362405124).
文摘We present a novel method for scale-invariant 3D face recognition by integrating computer-generated holography with the Mellin transform.This approach leverages the scale-invariance property of the Mellin transform to address challenges related to variations in 3D facial sizes during recognition.By applying the Mellin transform to computer-generated holograms and performing correlation between them,which,to the best of our knowledge,is being done for the first time,we have developed a robust recognition framework capable of managing significant scale variations without compromising recognition accuracy.Digital holograms of 3D faces are generated from a face database,and the Mellin transform is employed to enable robust recognition across scale factors ranging from 0.4 to 2.0.Within this range,the method achieves 100%recognition accuracy,as confirmed by both simulation-based and hybrid optical/digital experimental validations.Numerical calculations demonstrate that our method significantly enhances the accuracy and reliability of 3D face recognition,as evidenced by the sharp correlation peaks and higher peak-to-noise ratio(PNR)values than that of using conventional holograms without the Mellin transform.Additionally,the hybrid optical/digital joint transform correlation hardware further validates the method's effectiveness,demonstrating its capability to accurately identify and distinguish 3D faces at various scales.This work provides a promising solution for advanced biometric systems,especially for those which require 3D scale-invariant recognition.
基金supported by the Fundamental Research Funds for the Central Universities (YJSJ25013,ZYTS25235,ZYTS25229)the National Natural Science Foundation of China (62274132, 62204189,62004151,62274126)+3 种基金the National Key R&D Program of China (2021YFF0500504,2021YFF0500501)the Natural Science Basic Research Plan in Shaanxi Province of China (2024GX-YBXM514)the Young Talent Fund of Association for Science and Technology in Shaanxi,China (20220115)the Interdisciplinary Cultivation Program of Xidian University (21103240003)。
文摘Scalable fabrication of homogeneous perovskite films remains crucial for bridging the efficiency gap between lab-scale solar cells and commercial solar modules.To tackle this issue,we introduce NCyano-4-methyl-N-phenylbenzenesulfonamide(CMPS) additives into perovskite precursors,enabling slot-die coating of large-area modules under ambient conditions.CMPS suppresses colloidal aggregation and delays crystallization,yielding high-quality uniform films.Small-area devices(0.07 cm^(2) aperture area) incorporating CMPS exhibited a significant efficiency increase from 22.07 % to 24.58 %.Corresponding encapsulated devices maintained 85 % of their initial power conversion efficiency(PCE,average 23.56 %) after 1500 h of continuous maximum power point(MPP) tracking under one-sun illumination at 50-55℃.Furthermore,we demonstrate impressive efficiency of perovskite solar modules,achieving 20.57 %(52 cm^(2) aperture area) for 10 cm × 10 cm mini-modules and 17.02 %(260 cm^(2) aperture area) for 21 cm × 21 cm sub-modules,representing the state-of-the-art performance for solutionprocessed devices at these scales.
文摘Optimizing routing and resource allocation in decentralized unmanned aerial vehicle(UAV)networks remains challenging due to interference and rapidly changing topologies.The authors introduce a novel framework combining double deep Q-networks(DDQNs)and graph neural networks(GNNs)for joint routing and resource allocation.The framework uses GNNs to model the network topology and DDQNs to adaptively control routing and resource allocation,addressing interference and improving network performance.Simulation results show that the proposed approach outperforms traditional methods such as Closest-to-Destination(c2Dst),Max-SINR(mSINR),and Multi-Layer Perceptron(MLP)-based models,achieving approximately 23.5% improvement in throughput,50% increase in connection probability,and 17.6% reduction in number of hops,demonstrating its effectiveness in dynamic UAV networks.
文摘Federated Learning(FL)has become a leading decentralized solution that enables multiple clients to train a model in a collaborative environment without directly sharing raw data,making it suitable for privacy-sensitive applications such as healthcare,finance,and smart systems.As the field continues to evolve,the research field has become more complex and scattered,covering different system designs,training methods,and privacy techniques.This survey is organized around the three core challenges:how the data is distributed,how models are synchronized,and how to defend against attacks.It provides a structured and up-to-date review of FL research from 2023 to 2025,offering a unified taxonomy that categorizes works by data distribution(Horizontal FL,Vertical FL,Federated Transfer Learning,and Personalized FL),training synchronization(synchronous and asynchronous FL),optimization strategies,and threat models(data leakage and poisoning attacks).In particular,we summarize the latest contributions in Vertical FL frameworks for secure multi-party learning,communication-efficient Horizontal FL,and domain-adaptive Federated Transfer Learning.Furthermore,we examine synchronization techniques addressing system heterogeneity,including straggler mitigation in synchronous FL and staleness management in asynchronous FL.The survey covers security threats in FL,such as gradient inversion,membership inference,and poisoning attacks,as well as their defense strategies that include privacy-preserving aggregation and anomaly detection.The paper concludes by outlining unresolved issues and highlighting challenges in handling personalized models,scalability,and real-world adoption.
文摘In contemporary computer vision,convolutional neural networks(CNNs)and vision transformers(ViTs)represent the two primary architectural paradigms for image recognition.While both approaches have been widely adopted in medical imaging applications,they operate based on fundamentally different computational principles.This report attempts to provide brief application notes on ViTs and CNNs,particularly focusing on scenarios that guide the selection of one architecture over the other in practical medical implementations.Generally,CNNs rely on convolutional kernels,localized receptive fields,and weight sharing,enabling efficient hierarchical feature extraction.These properties contribute to strong performance in detecting spatially constrained patterns such as textures,edges,and anatomical boundaries,while maintaining relatively low computational requirements.ViTs,on the other hand,decompose images into smaller segments referred to as tokens and employ self-attention mechanisms to model relationships across the entire image.This global modeling capability allows ViTs to capture long-range dependencies that may be difficult for convolution-based architectures to learn.However,ViTs typically achieve optimal performance when trained on extremely large datasets or when supported by extensive pretraining,as their reduced inductive bias requires greater data exposure to learn robust representations.This report briefly examines the architectural structure,underlying mathematical foundations,and relative performance characteristics of CNNs and ViTs,drawing upon recent findings from contemporary research.Emphasis is placed on understanding how differences in data availability,computational resources,and task requirements influence model effectiveness across medical imaging domains.Most importantly,the report serves as a concise application guide for practitioners seeking informed implementation decisions between these two influential deep learning frameworks.
基金the Engineering and Physical Sciences Research Council(EPSRC)for funding the researchUK India Education Research Initiative(UKIERI)for funding support.
文摘This review provides an insightful and comprehensive exploration of the emerging 2D material borophene,both pristine and modified,emphasizing its unique attributes and potential for sustainable applications.Borophene’s distinctive properties include its anisotropic crystal structures that contribute to its exceptional mechanical and electronic properties.The material exhibits superior electrical and thermal conductivity,surpassing many other 2D materials.Borophene’s unique atomic spin arrangements further diversify its potential application for magnetism.Surface and interface engineering,through doping,functionalization,and synthesis of hybridized and nanocomposite borophene-based systems,is crucial for tailoring borophene’s properties to specific applications.This review aims to address this knowledge gap through a comprehensive and critical analysis of different synthetic and functionalisation methods,to enhance surface reactivity by increasing active sites through doping and surface modifications.These approaches optimize diffusion pathways improving accessibility for catalytic reactions,and tailor the electronic density to tune the optical and electronic behavior.Key applications explored include energy systems(batteries,supercapacitors,and hydrogen storage),catalysis for hydrogen and oxygen evolution reactions,sensors,and optoelectronics for advanced photonic devices.The key to all these applications relies on strategies to introduce heteroatoms for tuning electronic and catalytic properties,employ chemical modifications to enhance stability and leverage borophene’s conductivity and reactivity for advanced photonics.Finally,the review addresses challenges and proposes solutions such as encapsulation,functionalization,and integration with composites to mitigate oxidation sensitivity and overcome scalability barriers,enabling sustainable,commercial-scale applications.
基金support from the Roy A.Wilkens Professorship Endowment。
文摘Interferometry is a crucial investigative technique used across diverse fields to achieve highprecision measurements.It works by analyzing the phase difference between two interfering waves,which results from variations in optical path lengths within an interferometer.We introduce a novel method for directly measuring changes in the phase difference within an optical interferometer,importantly,with the added advantage of a controllable enhancement factor.This approach is achieved through a two-step process:first,the optical phase difference is encoded into a sub-GHz radiofrequency(RF)signal using microwave-photonic manipulation;then,RF interferometry-assisted phase amplification is implemented at the destructive interference point.In our experiments,we demonstrate a phase sensitivity of 2.14 rad∕nm operating at 140 MHz using a miniature in-fiber Fabry-Pérot interferometer for sub-nanometer displacement sensing,which reveals a sensitivity magnification factor of 258.6.With further refinement,we anticipate that even higher enhancement factors can be achieved,paving the way for the development of cost-effective,ultrasensitive interferometry-based instruments for high-precision optical measurements.
基金supported by the European Research Council(ERC)under Grant Agreement No.951424(Water-Futures)by the Republic of Cyprus through the Deputy Ministry of Research,Innovation and Digital Policy.
文摘In today's connected world,the generation of massive streaming data across diverse domains has become commonplace.In the presence of concept drift,class imbalance,label scarcity,and new class emergence,these challenges jointly degrade representation stability,bias learning toward outdated distributions,and reduce the resilience and reliability of detection in dynamic environments.This paper proposes a streaming classincremental learning(SCIL)framework to address these issues.The SCIL framework integrates an autoencoder(AE)with a multi-layer perceptron for multi-class prediction,employs a dual-loss strategy(classification and reconstruction)for prediction and new class detection,uses corrected pseudo-labels for online training,manages classes with queues,and applies oversampling to handle imbalance.The rationale behind the method's structure is elucidated through ablation studies,and a comprehensive experimental evaluation is performed using both real-world and synthetic datasets that feature class imbalance,incremental classes,and concept drifts.Our results demonstrate that SCIL outperforms strong baselines and state-of-the-art methods.In line with our commitment to Open Science,we make our code and datasets available to the community.
基金the School of Engineering and Built Environment at Anglia Ruskin University,UK,for the supportthe support of IRC-CSS and the Electrical Engineering Department,KFUPM,Saudi Arabia。
文摘We discuss recent progress in using machine-learning(ML)-enabled inverse design techniques applied to photonic devices and components.Specifically,we highlight the design of optical sources,including fiber and semiconductor lasers,as well as Raman and semiconductor optical amplifiers.Although inverse design approaches for optical detectors remain relatively underexplored,we examine optical layers,particularly metamaterial absorbers,as promising candidates for high-performance optical detection.In addition,we underscore advancements in inverse designing passive optical components,including beam splitters,gratings,and optical fibers.These optical blocks are fundamental in developing next-generation standalone optical communication systems and optical sensing networks,including integrated sensing and communication technologies.While categorizing various reported deep learning architectures across five paradigms,we offer a paradigm-based perspective that reveals how different ML techniques function within modern inverse design methods and enable fast,data-driven solutions that significantly reduce design time and computational demands compared with traditional optimization methods.
基金supported by Singapore RIE2025 Manufacturing,Trade and Connectivity Industry Alignment Fund-Pre-Positioning(IAF-PP)under Grant M24N2a0039 through WP2-Intelligent Switching Controlthe National Research Foundation Singapore under its AI Singapore Programme under Grant AISG4-GC-2023-007-1B.
文摘In this paper,we propose a fuzzy logic-based coded event-triggered control with self-adjustable prescribed performance(FL-CEC-SPP)to address the trade-off between control performance and communication efficiency in resource-constrained networked control systems.The method integrates a fuzzy-coded event-triggered controller into a coded control framework to dynamically adjust the triggering threshold,thereby reducing unnecessary transmissions while maintaining system stability.A self-adjustable prescribed performance constraint is also incorporated to ensure that the tracking error remains within predefined bounds under arbitrary initial conditions.Theoretical analyses and simulation comparisons show that the method proposed in this paper maintains good tracking performance and stability while reducing the communication burden,and has wide applications in resource-constrained network control systems.
基金supported by the Smart Grid-National Science and Technology Major Project(2025ZD0804500)the National Natural Science Foundation of China under Grant 52307232the Hunan Provincial Natural Science Foundation of China under Grant 2024JJ4055.
文摘The hybrid series-parallel microgrid attracts more attention by combining the advantages of both the series-stacked voltage and parallel-expanded capacity.Low-voltage distributed generations(DGs)are connected in series to form the intra-string,and then multiple strings are interconnected in parallel.For the existing control strategies,both intra-string and inter-string depend on the centralized or distributed control with high communication reliance.It has limited scalability and redundancy under abnormal conditions.Alternatively,in this study,an intra-string distributed and inter-string decentralized control framework is proposed.Within the string,a few DGs close to the AC bus are the leaders to get the string power information and the rest DGs are the followers to acquire the synchronization information through the droop-based distributed consistency.Specifically,the output of the entire string has the active power−angular frequency(ω-P)droop characteristic,and the decentralized control among strings can be autonomously guaranteed.Moreover,the secondary control is designed to realize multi-mode objectives,including on/off-grid mode switching,grid-connected power interactive management,and off-grid voltage quality regulation.As a result,the proposed method has the ability of plug-and-play capabilities,single-point failure redundancy,and seamless mode-switching.Experimental results are provided to verify the effectiveness of the proposed practical solution.
文摘Reinforcement learning(RL),as an important branch of machine learning,has recently achieved extensive attention and success in many applications.Its main idea is to enable agents to continuously learn to make optimal decisions by trying to maximize a reward function for their actions and interactions with the environment.However,making highquality decisions in complex and uncertain real-world scenarios is a challenging task.The interference and attacks in such scenarios tend to destroy the existing strategies.Maintaining RL's optimal performance in various cases and adapting to changing environments remains an important challenge.This article presents a comprehensive review of recent advancements in robust reinforcement learning(RRL),and analyzes them from the perspectives of challenges,methodologies,and applications.It systematically evaluates current progress in RRL and summarizes the commonly used benchmark platforms.Finally,several open challenges are discussed to stimulate further research and guide future developments in this area.
基金the Chinese University of Hong Kong for providing research resources and institutional support
文摘Accurate blood pressure(BP)monitoring is essential for preventing and managing cardiovascular disease.Advancements in materials science,medicine,flexible electronic,and artificial intelligence(AI)have enabled cuffless,unobtrusive BP monitoring systems,offering an alternative to traditional sphygmomanometers.However,extending these advances to real-world cardiovascular care particularly in resource-limited settings remains challenging due to constraints in computational resources,power efficiency,and deployment scalability.This review presents a comprehensive synthesis of AI-enhanced wearable BP monitoring,emphasizing its potential for personalized,scalable,and accessible healthcare.We systematically analyze the end-to-end system architecture,from mechano-electric sensing principles and AI-based estimation models to edge-aware deployment strategies tailored for low-resource environments.We further discuss clinical validation metrics and implementation barriers and prospective strategies.To bridge lab-to-field translation,we propose an innovative"sensor-model-deployment-assessment"co-design framework.This roadmap highlights how AI-enhanced BP technologies can support proactive hypertension control and promote cardiovascular health equity on a global scale.
基金funded by the Research,Development,and Innovation Authority(RDIA)—Kingdom of Saudi Arabia(Grant No.13292-psu-2023-PSNU-R-3-1-EF-).
文摘Colorectal cancer is the third most diagnosed cancer worldwide,and immune checkpoint inhibitors have shown promising therapeutic outcomes in selected patient groups.This study performed a comprehensive analysis of multi-omics data from The Cancer Genome Atlas colorectal adenocarcinoma cohort(TCGA-COADREAD),accessed through cBioPortal,to develop machine learning models for predicting progression-free survival(PFS)following immunotherapy.The dataset included clinical variables,genomic alterations in Kirsten Rat Sarcoma Viral Oncogene Homolog(KRAS),B-Raf Proto-Oncogene(BRAF),and Neuroblastoma RAS Viral Oncogene Homolog(NRAS),microsatellite instability(MSI)status,tumor mutation burden(TMB),and expression of immune checkpoint genes.Kaplan–Meier analysis showed that KRAS mutations were significantly associated with reduced PFS,while BRAF and NRAS mutations had no significant impact.MSI-high tumors exhibited elevated TMB and increased immune checkpoint expression,reflecting their immunologically active phenotype.We developed both survival and classification models,with the Extra Trees classifier achieving the best performance(accuracy=0.86,precision=0.67,recall=0.70,F1-score=0.68,AUC=0.84).These findings highlight the potential of combining genomic and immune biomarkers with machine learning to improve patient stratification and guide personalized immunotherapy decisions.An interactive web application was also developed to enable clinicians to input patient-specific molecular and clinical data and visualize individualized PFS predictions,supporting timely,data-driven treatment planning.
基金financially supported by the National Natural Science Foundation of China(52473178,52473275)Young Elite Scientists Sponsorship Program by CAST(2022QNRC001)+8 种基金the Program of Introducing Talents of Jiangnan University(1065219032210150)the Postgraduate Research&Practice Innovation Program of Jiangsu Province(KYCX23_2474)the Science and Technology Program of Jiangsu Administration for Market Regulation(KJ2024013)the Wuxi Science and Technology Development Fund Project(K20231042)Funded by Basic Research Program of Jiangsu(BK20251613)the Fundamental Research Funds for the Central Universities(JUSRP202504025)Jiangnan University Student Innovation Program(2025CXZ066)Laboratory of Flexible Electronics Technology,Tsinghua Universitythe Wuxi Taihu Talent Innovation Project(2024)。
文摘Textile electronics with extraordinary sensing capabilities holds significant potential in the Artificial Intelligence of Things(AIoT).However,little effort is paid to their mutual advantages of robust interfacial interactions,ultra-strong mechanical performance,and stability.Herein,we fabricate homogeneous and multifunctional core-shell macrofibers by integrating bridge-functionalized MXene/PEDOT:PSS conductive ink with aligned bacterial cellulose(BC).These resulting macrofibers feature mechanical properties(tensile strength of 433.2 MPa and the Young's modulus of 25.9 GPa),exceptional electrical conductivity(10.05 S cm^(-1))and durable hydrophobicity.Such superior robustness allows for the fabrication of the macrofibers woven into textile-based triboelectric nanogenerator(PKT-TENG)and shows an impressive high-performance of a maximum open-circuit voltage of 272.54 V,short-circuit current of 14.56μA and power density of 86.29 mW m^(-2),which successfully powers commercial electronics.As the proof-of-concept illustration,the macrofibers with durable hydrophobicity and high piezoresistive sensitivity are further employed for precepting diverse liquids that can simultaneously monitor their distinctive motion features via real-time resistance variation on the textile-based array.This work is expected to offer new insights into the design of advanced fibers with ultra-strong mechanical capabilities and high conductivity and provide an avenue for the development of textile electronics for high-performance sensing and intelligent manufacturing.
基金supported by the Research Projects of the Department of Education of Guangdong Province 2024ZDZX3079The financial support from the Guangdong Basic and Applied Basic Research Foundation(No.2023A1515011677)+4 种基金the Scientific and Technical Innovation Council of Shenzhen(20220812165832002)the Research Projects of Department of Education of Guangdong Province-2023GCZX015the Innovation Team Project of Guangdong(2022KCXTD055)the China Postdoctoral Science Foundation(Certificate Number:2024M763441)is gratefully acknowledgedsupported by the Postdoctoral Fellowship Program of CPSF under Grant Number GZB20250031 and Research Projects of the Department of Education of Guangdong Province 2023GCZX015。
文摘Despite the intrinsic durability of polymeric hole transport materials,poly-triarylamines(PTAA)-based inverted perovskite solar cells(PSCs)have lagged behind their counterparts in efficiency,primarily due to poor surface wettability,insufficient interfacial contact,and unfavorable energy level alignment at the PTAA/perovskite interface.Here,we report a highly effective interfacial engineering strategy employing the ionic liquid 1,3-dimethylimidazolium dimethyl phosphate(DMIMPH)as a multifunctional interfacial modifier.The incorporation of DMIMPH improves PTAA wettability,promoting the growth of high-quality perovskite films with enhanced interfacial contact.Concurrently,DMIMPH effectively tunes the energy levels of PTAA,enhances its electrical conductivity,and passivates interfacial defects with more efficient hole extraction and charge transport.Moreover,its interaction with residual PbI_(2) modulates perovskite crystallization kinetics,yielding highly crystalline perovskite films with enlarged grain sizes,reduced PbI_(2) residue,and suppressed trap densities.As a result,PTAA-based p-i-n PSCs employing this approach achieve a record certified power conversion efficiency(PCE)of 24.52%,with a champion efficiency of 25.12%—the highest certified value for PTAA-based perovskite devices to date.Impressively,the DMIMPH-modified PSCs without encapsulation maintained 87.48%of their initial efficiency after 1600 h in air.This strategy offers an effective pathway for advancing the performance and stability of polymer-based inverted PSCs.
文摘Over the past years,many efforts have been accomplished to achieve fast and accurate meta-heuristic algorithms to optimize a variety of real-world problems.This study presents a new optimization method based on an unusual geological phenomenon in nature,named Geyser inspired Algorithm(GEA).The mathematical modeling of this geological phenomenon is carried out to have a better understanding of the optimization process.The efficiency and accuracy of GEA are verified using statistical examination and convergence rate comparison on numerous CEC 2005,CEC 2014,CEC 2017,and real-parameter benchmark functions.Moreover,GEA has been applied to several real-parameter engineering optimization problems to evaluate its effectiveness.In addition,to demonstrate the applicability and robustness of GEA,a comprehensive investigation is performed for a fair comparison with other standard optimization methods.The results demonstrate that GEA is noticeably prosperous in reaching the optimal solutions with a high convergence rate in comparison with other well-known nature-inspired algorithms,including ABC,BBO,PSO,and RCGA.Note that the source code of the GEA is publicly available at https://www.optim-app.com/projects/gea.
基金financially supported via Australian Research Council(FT180100705)the support by the National Natural Science Foundation of China(22209103)+3 种基金the support from UTS Chancellor's Research Fellowshipsthe support from Open Project of State Key Laboratory of Advanced Special Steel,the Shanghai Key Laboratory of Advanced Ferrometallurgy,Shanghai University(SKLASS 2021-**)Joint International Laboratory on Environmental and Energy Frontier MaterialsInnovation Research Team of High-Level Local Universities in Shanghai。
文摘Electrochemical carbon dioxide reduction reaction(CO_(2)RR)provides a promising way to convert CO_(2)to chemicals.The multicarbon(C_(2+))products,especially ethylene,are of great interest due to their versatile industrial applications.However,selectively reducing CO_(2)to ethylene is still challenging as the additional energy required for the C–C coupling step results in large overpotential and many competing products.Nonetheless,mechanistic understanding of the key steps and preferred reaction pathways/conditions,as well as rational design of novel catalysts for ethylene production have been regarded as promising approaches to achieving the highly efficient and selective CO_(2)RR.In this review,we first illustrate the key steps for CO_(2)RR to ethylene(e.g.,CO_(2)adsorption/activation,formation of~*CO intermediate,C–C coupling step),offering mechanistic understanding of CO_(2)RR conversion to ethylene.Then the alternative reaction pathways and conditions for the formation of ethylene and competitive products(C_1 and other C_(2+)products)are investigated,guiding the further design and development of preferred conditions for ethylene generation.Engineering strategies of Cu-based catalysts for CO_(2)RR-ethylene are further summarized,and the correlations of reaction mechanism/pathways,engineering strategies and selectivity are elaborated.Finally,major challenges and perspectives in the research area of CO_(2)RR are proposed for future development and practical applications.
基金the National Natural Science Foundation of China(62271485,61903363,U1811463,62103411,62203250)the Science and Technology Development Fund of Macao SAR(0093/2023/RIA2,0050/2020/A1)。
文摘DURING our discussion at workshops for writing“What Does ChatGPT Say:The DAO from Algorithmic Intelligence to Linguistic Intelligence”[1],we had expected the next milestone for Artificial Intelligence(AI)would be in the direction of Imaginative Intelligence(II),i.e.,something similar to automatic wordsto-videos generation or intelligent digital movies/theater technology that could be used for conducting new“Artificiofactual Experiments”[2]to replace conventional“Counterfactual Experiments”in scientific research and technical development for both natural and social studies[2]-[6].Now we have OpenAI’s Sora,so soon,but this is not the final,actually far away,and it is just the beginning.
