With the advent of modern devices,such as smartphones and wearable devices,high-dimensional data are collected on many participants for a period of time or even in perpetuity.For this type of data,dependencies between...With the advent of modern devices,such as smartphones and wearable devices,high-dimensional data are collected on many participants for a period of time or even in perpetuity.For this type of data,dependencies between and within data batches exist because data are collected from the same individual over time.Under the framework of streamed data,individual historical data are not available due to the storage and computation burden.It is urgent to develop computationally efficient methods with statistical guarantees to analyze high-dimensional streamed data and make reliable inferences in practice.In addition,the homogeneity assumption on the model parameters may not be valid in practice over time.To address the above issues,in this paper,we develop a new renewable debiased-lasso inference method for high-dimensional streamed data allowing dependences between and within data batches to exist and model parameters to gradually change.We establish the large sample properties of the proposed estimators,including consistency and asymptotic normality.The numerical results,including simulations and real data analysis,show the superior performance of the proposed method.展开更多
Groundwater and surface water contamination by PTE(Potentially Toxic Elements)was assessed in Ruashi and Annexe municipalities of Lubumbashi city.Analyses of seventy water samples collected from six drilled wells,eigh...Groundwater and surface water contamination by PTE(Potentially Toxic Elements)was assessed in Ruashi and Annexe municipalities of Lubumbashi city.Analyses of seventy water samples collected from six drilled wells,eight spade-sunk wells,one river and one spring in both municipalities in 2017 and 2018 were carried out by ICP-SF-MS(Inductively Coupled Plasma-Sector Field Mass Spectrometry).Twenty PTEs including aluminum,arsenic,barium,bismuth,cadmium,cesium,chromium,cobalt,copper,iron,lead,manganese,molybdenum,nickel,strontium,thallium,tungsten,uranium,vanadium and zinc were detected at various concentrations in each one of the samples.Many samples had concentrations and mean concentrations of PTEs,such as aluminum,cadmium,copper,iron,lead,manganese,nickel and zinc,higher than the respective acceptable limits set for drinking water by the EU(European Union),the USEPA(United States Environmental Protection Agency),and the WHO(World Health Organization)standards.Most PTEs being deleterious to human health even at very low concentrations,people who use the groundwater and surface water to meet their water needs in both Ruashi and Annexe municipalities are at risk.展开更多
Depleting fossil energy sources and conventional polluting power generation pose a threat to sustainable development.Hydroelectricity generation from ubiquitous and spontaneous phase transitions between liquid and gas...Depleting fossil energy sources and conventional polluting power generation pose a threat to sustainable development.Hydroelectricity generation from ubiquitous and spontaneous phase transitions between liquid and gaseous water has been considered a promising strategy for mitigating the energy crisis.Fibrous materials with unique flexibility,processability,multifunctionality,and practicability have been widely applied for fibrous materials-based hydroelectricity generation(FHG).In this review,the power generation mechanisms,design principles,and electricity enhancement factors of FHG are first introduced.Then,the fabrication strategies and characteristics of varied constructions including 1D fiber,1D yarn,2D fabric,2D membrane,3D fibrous framework,and 3D fibrous gel are demonstrated.Afterward,the advanced functions of FHG during water harvesting,proton dissociation,ion separation,and charge accumulation processes are analyzed in detail.Moreover,the potential applications including power supply,energy storage,electrical sensor,and information expression are also discussed.Finally,some existing challenges are considered and prospects for future development are sincerely proposed.展开更多
Micro-expressions(ME)recognition is a complex task that requires advanced techniques to extract informative features fromfacial expressions.Numerous deep neural networks(DNNs)with convolutional structures have been pr...Micro-expressions(ME)recognition is a complex task that requires advanced techniques to extract informative features fromfacial expressions.Numerous deep neural networks(DNNs)with convolutional structures have been proposed.However,unlike DNNs,shallow convolutional neural networks often outperform deeper models in mitigating overfitting,particularly with small datasets.Still,many of these methods rely on a single feature for recognition,resulting in an insufficient ability to extract highly effective features.To address this limitation,in this paper,an Improved Dual-stream Shallow Convolutional Neural Network based on an Extreme Gradient Boosting Algorithm(IDSSCNN-XgBoost)is introduced for ME Recognition.The proposed method utilizes a dual-stream architecture where motion vectors(temporal features)are extracted using Optical Flow TV-L1 and amplify subtle changes(spatial features)via EulerianVideoMagnification(EVM).These features are processed by IDSSCNN,with an attention mechanism applied to refine the extracted effective features.The outputs are then fused,concatenated,and classified using the XgBoost algorithm.This comprehensive approach significantly improves recognition accuracy by leveraging the strengths of both temporal and spatial information,supported by the robust classification power of XgBoost.The proposed method is evaluated on three publicly available ME databases named Chinese Academy of Sciences Micro-expression Database(CASMEII),Spontaneous Micro-Expression Database(SMICHS),and Spontaneous Actions and Micro-Movements(SAMM).Experimental results indicate that the proposed model can achieve outstanding results compared to recent models.The accuracy results are 79.01%,69.22%,and 68.99%on CASMEII,SMIC-HS,and SAMM,and the F1-score are 75.47%,68.91%,and 63.84%,respectively.The proposed method has the advantage of operational efficiency and less computational time.展开更多
The dorsal and ventral visual streams have been considered to play distinct roles in visual processing for action:the dorsal stream is assumed to support real-time actions,while the ventral stream facilitates memory-g...The dorsal and ventral visual streams have been considered to play distinct roles in visual processing for action:the dorsal stream is assumed to support real-time actions,while the ventral stream facilitates memory-guided actions.However,recent evidence suggests a more integrated function of these streams.We investigated the neural dynamics and functional connectivity between them during memory-guided actions using intracranial EEG.We tracked neural activity in the inferior parietal lobule in the dorsal stream,and the ventral temporal cortex in the ventral stream as well as the hippocampus during a delayed action task involving object identity and location memory.We found increased alpha power in both streams during the delay,indicating their role in maintaining spatial visual information.In addition,we recorded increased alpha power in the hippocampus during the delay,but only when both object identity and location needed to be remembered.We also recorded an increase in theta band phase synchronization between the inferior parietal lobule and ventral temporal cortex and between the inferior parietal lobule and hippocampus during the encoding and delay.Granger causality analysis indicated dynamic and frequency-specific directional interactions among the inferior parietal lobule,ventral temporal cortex,and hippocampus that varied across task phases.Our study provides unique electrophysiological evidence for close interactions between dorsal and ventral streams,supporting an integrated processing model in which both streams contribute to memory-guided actions.展开更多
Sign language dataset is essential in sign language recognition and translation(SLRT). Current public sign language datasets are small and lack diversity, which does not meet the practical application requirements for...Sign language dataset is essential in sign language recognition and translation(SLRT). Current public sign language datasets are small and lack diversity, which does not meet the practical application requirements for SLRT. However, making a large-scale and diverse sign language dataset is difficult as sign language data on the Internet is scarce. In making a large-scale and diverse sign language dataset, some sign language data qualities are not up to standard. This paper proposes a two information streams transformer(TIST) model to judge whether the quality of sign language data is qualified. To verify that TIST effectively improves sign language recognition(SLR), we make two datasets, the screened dataset and the unscreened dataset. In this experiment, this paper uses visual alignment constraint(VAC) as the baseline model. The experimental results show that the screened dataset can achieve better word error rate(WER) than the unscreened dataset.展开更多
The exponential expansion of the Internet of Things(IoT),Industrial Internet of Things(IIoT),and Transportation Management of Things(TMoT)produces vast amounts of real-time streaming data.Ensuring system dependability...The exponential expansion of the Internet of Things(IoT),Industrial Internet of Things(IIoT),and Transportation Management of Things(TMoT)produces vast amounts of real-time streaming data.Ensuring system dependability,operational efficiency,and security depends on the identification of anomalies in these dynamic and resource-constrained systems.Due to their high computational requirements and inability to efficiently process continuous data streams,traditional anomaly detection techniques often fail in IoT systems.This work presents a resource-efficient adaptive anomaly detection model for real-time streaming data in IoT systems.Extensive experiments were carried out on multiple real-world datasets,achieving an average accuracy score of 96.06%with an execution time close to 7.5 milliseconds for each individual streaming data point,demonstrating its potential for real-time,resourceconstrained applications.The model uses Principal Component Analysis(PCA)for dimensionality reduction and a Z-score technique for anomaly detection.It maintains a low computational footprint with a sliding window mechanism,enabling incremental data processing and identification of both transient and sustained anomalies without storing historical data.The system uses a Multivariate Linear Regression(MLR)based imputation technique that estimates missing or corrupted sensor values,preserving data integrity prior to anomaly detection.The suggested solution is appropriate for many uses in smart cities,industrial automation,environmental monitoring,IoT security,and intelligent transportation systems,and is particularly well-suited for resource-constrained edge devices.展开更多
While feeding her ostriches,Wang Xue swiftly responds to fans'questions with a sense of humor through her popular live‑streaming channel.Since Wang,30,an expert in ostrich farming,started live broadcasting on a sh...While feeding her ostriches,Wang Xue swiftly responds to fans'questions with a sense of humor through her popular live‑streaming channel.Since Wang,30,an expert in ostrich farming,started live broadcasting on a short‑video platform with her husband Han Peng,their ostrich farming business has flourished,garnering nearly 3 million fans on the platform.“The highest viewership for one single live broadcast reached over 6 million,”said Wang,adding that riding an ostrich at high speed was an essential part of her broadcasts.Through her videos,Wang brought fans close to the ostriches,providing details of their habits,farming methods and economic value.展开更多
With the widespread application of Internet of Things(IoT)technology,the processing of massive realtime streaming data poses significant challenges to the computational and data-processing capabilities of systems.Alth...With the widespread application of Internet of Things(IoT)technology,the processing of massive realtime streaming data poses significant challenges to the computational and data-processing capabilities of systems.Although distributed streaming data processing frameworks such asApache Flink andApache Spark Streaming provide solutions,meeting stringent response time requirements while ensuring high throughput and resource utilization remains an urgent problem.To address this,the study proposes a formal modeling approach based on Performance Evaluation Process Algebra(PEPA),which abstracts the core components and interactions of cloud-based distributed streaming data processing systems.Additionally,a generic service flow generation algorithmis introduced,enabling the automatic extraction of service flows fromthe PEPAmodel and the computation of key performance metrics,including response time,throughput,and resource utilization.The novelty of this work lies in the integration of PEPA-based formal modeling with the service flow generation algorithm,bridging the gap between formal modeling and practical performance evaluation for IoT systems.Simulation experiments demonstrate that optimizing the execution efficiency of components can significantly improve system performance.For instance,increasing the task execution rate from 10 to 100 improves system performance by 9.53%,while further increasing it to 200 results in a 21.58%improvement.However,diminishing returns are observed when the execution rate reaches 500,with only a 0.42%gain.Similarly,increasing the number of TaskManagers from 10 to 20 improves response time by 18.49%,but the improvement slows to 6.06% when increasing from 20 to 50,highlighting the importance of co-optimizing component efficiency and resource management to achieve substantial performance gains.This study provides a systematic framework for analyzing and optimizing the performance of IoT systems for large-scale real-time streaming data processing.The proposed approach not only identifies performance bottlenecks but also offers insights into improving system efficiency under different configurations and workloads.展开更多
Real-time surveillance is attributed to recognizing the variety of actions performed by humans.Human Action Recognition(HAR)is a technique that recognizes human actions from a video stream.A range of variations in hum...Real-time surveillance is attributed to recognizing the variety of actions performed by humans.Human Action Recognition(HAR)is a technique that recognizes human actions from a video stream.A range of variations in human actions makes it difficult to recognize with considerable accuracy.This paper presents a novel deep neural network architecture called Attention RB-Net for HAR using video frames.The input is provided to the model in the form of video frames.The proposed deep architecture is based on the unique structuring of residual blocks with several filter sizes.Features are extracted from each frame via several operations with specific parameters defined in the presented novel Attention-based Residual Bottleneck(Attention-RB)DCNN architecture.A fully connected layer receives an attention-based features matrix,and final classification is performed.Several hyperparameters of the proposed model are initialized using Bayesian Optimization(BO)and later utilized in the trained model for testing.In testing,features are extracted from the self-attention layer and passed to neural network classifiers for the final action classification.Two highly cited datasets,HMDB51 and UCF101,were used to validate the proposed architecture and obtained an average accuracy of 87.70%and 97.30%,respectively.The deep convolutional neural network(DCNN)architecture is compared with state-of-the-art(SOTA)methods,including pre-trained models,inside blocks,and recently published techniques,and performs better.展开更多
No-wash bioassays based on nanoparticles are used widely in biochemical procedures because of their responsive sensing and no need forwashing processes.Essential for no-wash biosensing are the interactions between nan...No-wash bioassays based on nanoparticles are used widely in biochemical procedures because of their responsive sensing and no need forwashing processes.Essential for no-wash biosensing are the interactions between nanoparticles and biomolecules,but it is challenging toachieve controlled bioconjugation of molecules on nanomaterials.Reported here is a way to actively improve nanoparticle-based no-washbioassays by enhancing the binding between biomolecules and gold nanoparticles via acoustic streaming generated by a gigahertz piezoelectricnanoelectromechanical resonator.Tunable micro-vortices are generated at the device-liquid interface,thereby accelerating the internalcirculating flow of the solution,bypassing the diffusion limitation,and thus improving the binding between the biomolecules and goldnanoparticles.Combined with fluorescence quenching,an enhanced and ultrafast no-wash biosensing assay is realized for specific proteins.The sensing method presented here is a versatile tool for different types of biomolecule detection with high efficiency and simplicity.展开更多
Many real⁃world machine learning applications face the challenge of dealing with changing data over time,known as concept drift,and the issue of data indeterminacy,where all the true labels available are unrealistic.T...Many real⁃world machine learning applications face the challenge of dealing with changing data over time,known as concept drift,and the issue of data indeterminacy,where all the true labels available are unrealistic.This can lead to a decrease in the accuracy of the prediction models.The aim of this study is to introduce a new approach for detecting drift,which is based on neutrosophic set theory.This approach takes into account uncertainty in the prediction model and is able to handle indeterminate information,considering its impact on the models performance.The proposed method reads data into windows and calculates a set of values based on the concept of neutrosophic membership.These values are then used in the Neutrosophic Support Vector Machine(N⁃SVM).To address the issue of indeterminate true label data,the values issued by N⁃SVM are expressed as entropy and used as input for the ADWIN(Adaptive Windowing)change detector.When a drift is detected,the prediction model is retrained by including only the most recent instances with the original training data set.The proposed method gives promising results in terms of drift detection accuracy compared to the state of existing drift detection methods such as KSWIN,ADWIN,and DWM.展开更多
With technological advancements,high-speed rail has emerged as a prevalent mode of transportation.During travel,passengers exhibit a growing demand for streaming media services.However,the high-speed mobile networks e...With technological advancements,high-speed rail has emerged as a prevalent mode of transportation.During travel,passengers exhibit a growing demand for streaming media services.However,the high-speed mobile networks environment poses challenges,including frequent base station handoffs,which significantly degrade wireless network transmission performance.Improving transmission efficiency in high-speed mobile networks and optimizing spatiotemporal wireless resource allocation to enhance passengers’media experiences are key research priorities.To address these issues,we propose an Adaptive Cross-Layer Optimization Transmission Method with Environment Awareness(ACOTM-EA)tailored for high-speed rail streaming media.Within this framework,we develop a channel quality prediction model utilizing Kalman filtering and an algorithm to identify packet loss causes.Additionally,we introduce a proactive base station handoffstrategy to minimize handoffrelated disruptions and optimize resource distribution across adjacent base stations.Moreover,this study presents a wireless resource allocation approach based on an enhanced genetic algorithm,coupled with an adaptive bitrate selection mechanism,to maximize passenger Quality of Experience(QoE).To evaluate the proposed method,we designed a simulation experiment and compared ACOTM-EA with established algorithms.Results indicate that ACOTM-EA improves throughput by 11%and enhances passengers’media experience by 5%.展开更多
In the era of big data,the growing number of real-time data streams often contains a lot of sensitive privacy information.Releasing or sharing this data directly without processing will lead to serious privacy informa...In the era of big data,the growing number of real-time data streams often contains a lot of sensitive privacy information.Releasing or sharing this data directly without processing will lead to serious privacy information leakage.This poses a great challenge to conventional privacy protection mechanisms(CPPM).The existing data partitioning methods ignore the number of data replications and information exchanges,resulting in complex distance calculations and inefficient indexing for high-dimensional data.Therefore,CPPM often fails to meet the stringent requirements of efficiency and reliability,especially in dynamic spatiotemporal environments.Addressing this concern,we proposed the Principal Component Enhanced Vantage-point tree(PEV-Tree),which is an enhanced data structure based on the idea of dimension reduction,and constructed a Distributed Spatio-Temporal Privacy Preservation Mechanism(DST-PPM)on it.In this work,principal component analysis and the vantage tree are used to establish the PEV-Tree.In addition,we designed three distributed anonymization algorithms for data streams.These algorithms are named CK-AA,CL-DA,and CT-CA,fulfill the anonymization rules of K-Anonymity,L-Diversity,and T-Closeness,respectively,which have different computational complexities and reliabilities.The higher the complexity,the lower the risk of privacy leakage.DST-PPM can reduce the dimension of high-dimensional information while preserving data characteristics and dividing the data space into vantage points based on distance.It effectively enhances the data processing workflow and increases algorithmefficiency.To verify the validity of the method in this paper,we conducted empirical tests of CK-AA,CL-DA,and CT-CA on conventional datasets and the PEV-Tree,respectively.Based on the big data background of the Internet of Vehicles,we conducted experiments using artificial simulated on-board network data.The results demonstrated that the operational efficiency of the CK-AA,CL-DA,and CT-CA is enhanced by 15.12%,24.55%,and 52.74%,respectively,when deployed on the PEV-Tree.Simultaneously,during homogeneity attacks,the probabilities of information leakage were reduced by 2.31%,1.76%,and 0.19%,respectively.Furthermore,these algorithms showcased superior utility(scalability)when executed across PEV-Trees of varying scales in comparison to their performance on conventional data structures.It indicates that DST-PPM offers marked advantages over CPPM in terms of efficiency,reliability,and scalability.展开更多
Unmanned Aerial Vehicles(UAVs)coupled with deep learning such as Convolutional Neural Networks(CNNs)have been widely applied across numerous domains,including agriculture,smart city monitoring,and fire rescue operatio...Unmanned Aerial Vehicles(UAVs)coupled with deep learning such as Convolutional Neural Networks(CNNs)have been widely applied across numerous domains,including agriculture,smart city monitoring,and fire rescue operations,owing to their malleability and versatility.However,the computation-intensive and latency-sensitive natures of CNNs present a formidable obstacle to their deployment on resource-constrained UAVs.Some early studies have explored a hybrid approach that dynamically switches between lightweight and complex models to balance accuracy and latency.However,they often overlook scenarios involving multiple concurrent CNN streams,where competition for resources between streams can substantially impact latency and overall system performance.In this paper,we first investigate the deployment of both lightweight and complex models for multiple CNN streams in UAV swarm.Specifically,we formulate an optimization problem to minimize the total latency across multiple CNN streams,under the constraints on UAV memory and the accuracy requirement of each stream.To address this problem,we propose an algorithm called Adaptive Model Switching of collaborative inference for MultiCNN streams(AMSM)to identify the inference strategy with a low latency.Simulation results demonstrate that the proposed AMSM algorithm consistently achieves the lowest latency while meeting the accuracy requirements compared to benchmark algorithms.展开更多
On May 14th,following the U.S.adjustment of additional tariffs on Chinese goods,American buyers began stockpiling in earnest.Many cross-border e-commerce companies also received a surge of orders.At 7 PM,a bustling Ha...On May 14th,following the U.S.adjustment of additional tariffs on Chinese goods,American buyers began stockpiling in earnest.Many cross-border e-commerce companies also received a surge of orders.At 7 PM,a bustling Hangzhou-based cross-border e-commerce company was alive with multiple languages echoing through its live-streaming rooms as backend order numbers climbed steadily.展开更多
In the contemporary digital landscape,the proliferation of information has led to an increasing diversity of channels through which consumers obtain information,resulting in a gradual transformation of shopping habits...In the contemporary digital landscape,the proliferation of information has led to an increasing diversity of channels through which consumers obtain information,resulting in a gradual transformation of shopping habits.Consumers now frequently rely on external sources to make well-informed purchasing decisions,leading to the emergence of live shopping as a prominent avenue for gathering product information and completing transactions.E-commerce live streaming has experienced rapid growth,leveraging its ability to generate traffic and capture consumer attention.The integration of content and live streaming not only meets users’psychological needs but also facilitates seamless communication between buyers and sellers.From the perspective of content marketing typologies,this paper examines content marketing across three key dimensions:informational content,entertainment content,and emotional content.It further explores the impact of content marketing on consumers’purchase intentions within the context of e-commerce live streaming.展开更多
基金Supported by National Key R&D Program of China(Grant No.2022YFA1003702)National Natural Science Foundation of China(Grant No.12271441)。
文摘With the advent of modern devices,such as smartphones and wearable devices,high-dimensional data are collected on many participants for a period of time or even in perpetuity.For this type of data,dependencies between and within data batches exist because data are collected from the same individual over time.Under the framework of streamed data,individual historical data are not available due to the storage and computation burden.It is urgent to develop computationally efficient methods with statistical guarantees to analyze high-dimensional streamed data and make reliable inferences in practice.In addition,the homogeneity assumption on the model parameters may not be valid in practice over time.To address the above issues,in this paper,we develop a new renewable debiased-lasso inference method for high-dimensional streamed data allowing dependences between and within data batches to exist and model parameters to gradually change.We establish the large sample properties of the proposed estimators,including consistency and asymptotic normality.The numerical results,including simulations and real data analysis,show the superior performance of the proposed method.
基金funding received from UNESCO-SIDA Project as well as Professor Martine Leermakers and Professor Willy Baeyens for their financial help to analyze the water samples in their laboratory at VUB.Acknowledgements
文摘Groundwater and surface water contamination by PTE(Potentially Toxic Elements)was assessed in Ruashi and Annexe municipalities of Lubumbashi city.Analyses of seventy water samples collected from six drilled wells,eight spade-sunk wells,one river and one spring in both municipalities in 2017 and 2018 were carried out by ICP-SF-MS(Inductively Coupled Plasma-Sector Field Mass Spectrometry).Twenty PTEs including aluminum,arsenic,barium,bismuth,cadmium,cesium,chromium,cobalt,copper,iron,lead,manganese,molybdenum,nickel,strontium,thallium,tungsten,uranium,vanadium and zinc were detected at various concentrations in each one of the samples.Many samples had concentrations and mean concentrations of PTEs,such as aluminum,cadmium,copper,iron,lead,manganese,nickel and zinc,higher than the respective acceptable limits set for drinking water by the EU(European Union),the USEPA(United States Environmental Protection Agency),and the WHO(World Health Organization)standards.Most PTEs being deleterious to human health even at very low concentrations,people who use the groundwater and surface water to meet their water needs in both Ruashi and Annexe municipalities are at risk.
基金funding support from the National Key Research and Development Program of China(No.2022YFB3805800)the National Natural Science Foundation of China(52173059)+1 种基金The Major Basic Research Project of the Natural Science Foundation of the Jiangsu Higher Education Institutions(21KJA540002)Jiangsu Funding Program for Excellent Postdoctoral Talent(2022ZB555).
文摘Depleting fossil energy sources and conventional polluting power generation pose a threat to sustainable development.Hydroelectricity generation from ubiquitous and spontaneous phase transitions between liquid and gaseous water has been considered a promising strategy for mitigating the energy crisis.Fibrous materials with unique flexibility,processability,multifunctionality,and practicability have been widely applied for fibrous materials-based hydroelectricity generation(FHG).In this review,the power generation mechanisms,design principles,and electricity enhancement factors of FHG are first introduced.Then,the fabrication strategies and characteristics of varied constructions including 1D fiber,1D yarn,2D fabric,2D membrane,3D fibrous framework,and 3D fibrous gel are demonstrated.Afterward,the advanced functions of FHG during water harvesting,proton dissociation,ion separation,and charge accumulation processes are analyzed in detail.Moreover,the potential applications including power supply,energy storage,electrical sensor,and information expression are also discussed.Finally,some existing challenges are considered and prospects for future development are sincerely proposed.
基金supported by the Key Research and Development Program of Jiangsu Province under Grant BE2022059-3,CTBC Bank through the Industry-Academia Cooperation Project,as well as by the Ministry of Science and Technology of Taiwan through Grants MOST-108-2218-E-002-055,MOST-109-2223-E-009-002-MY3,MOST-109-2218-E-009-025,and MOST431109-2218-E-002-015.
文摘Micro-expressions(ME)recognition is a complex task that requires advanced techniques to extract informative features fromfacial expressions.Numerous deep neural networks(DNNs)with convolutional structures have been proposed.However,unlike DNNs,shallow convolutional neural networks often outperform deeper models in mitigating overfitting,particularly with small datasets.Still,many of these methods rely on a single feature for recognition,resulting in an insufficient ability to extract highly effective features.To address this limitation,in this paper,an Improved Dual-stream Shallow Convolutional Neural Network based on an Extreme Gradient Boosting Algorithm(IDSSCNN-XgBoost)is introduced for ME Recognition.The proposed method utilizes a dual-stream architecture where motion vectors(temporal features)are extracted using Optical Flow TV-L1 and amplify subtle changes(spatial features)via EulerianVideoMagnification(EVM).These features are processed by IDSSCNN,with an attention mechanism applied to refine the extracted effective features.The outputs are then fused,concatenated,and classified using the XgBoost algorithm.This comprehensive approach significantly improves recognition accuracy by leveraging the strengths of both temporal and spatial information,supported by the robust classification power of XgBoost.The proposed method is evaluated on three publicly available ME databases named Chinese Academy of Sciences Micro-expression Database(CASMEII),Spontaneous Micro-Expression Database(SMICHS),and Spontaneous Actions and Micro-Movements(SAMM).Experimental results indicate that the proposed model can achieve outstanding results compared to recent models.The accuracy results are 79.01%,69.22%,and 68.99%on CASMEII,SMIC-HS,and SAMM,and the F1-score are 75.47%,68.91%,and 63.84%,respectively.The proposed method has the advantage of operational efficiency and less computational time.
基金supported by European Union–Next Generation EU(LX22NPO5107(MEYS))the Czech Science Foundation(20-21339S)+2 种基金the Grant Agency of Charles University(GAUK 248122 and 272221)ERDF-Project Brain Dynamics(CZ.02.01.01/00/22_008/0004643)the Ministry of Health of the Czech Republic Project NU21J-08-00081.
文摘The dorsal and ventral visual streams have been considered to play distinct roles in visual processing for action:the dorsal stream is assumed to support real-time actions,while the ventral stream facilitates memory-guided actions.However,recent evidence suggests a more integrated function of these streams.We investigated the neural dynamics and functional connectivity between them during memory-guided actions using intracranial EEG.We tracked neural activity in the inferior parietal lobule in the dorsal stream,and the ventral temporal cortex in the ventral stream as well as the hippocampus during a delayed action task involving object identity and location memory.We found increased alpha power in both streams during the delay,indicating their role in maintaining spatial visual information.In addition,we recorded increased alpha power in the hippocampus during the delay,but only when both object identity and location needed to be remembered.We also recorded an increase in theta band phase synchronization between the inferior parietal lobule and ventral temporal cortex and between the inferior parietal lobule and hippocampus during the encoding and delay.Granger causality analysis indicated dynamic and frequency-specific directional interactions among the inferior parietal lobule,ventral temporal cortex,and hippocampus that varied across task phases.Our study provides unique electrophysiological evidence for close interactions between dorsal and ventral streams,supporting an integrated processing model in which both streams contribute to memory-guided actions.
基金supported by the National Language Commission to research on sign language data specifications for artificial intelligence applications and test standards for language service translation systems (No.ZDI145-70)。
文摘Sign language dataset is essential in sign language recognition and translation(SLRT). Current public sign language datasets are small and lack diversity, which does not meet the practical application requirements for SLRT. However, making a large-scale and diverse sign language dataset is difficult as sign language data on the Internet is scarce. In making a large-scale and diverse sign language dataset, some sign language data qualities are not up to standard. This paper proposes a two information streams transformer(TIST) model to judge whether the quality of sign language data is qualified. To verify that TIST effectively improves sign language recognition(SLR), we make two datasets, the screened dataset and the unscreened dataset. In this experiment, this paper uses visual alignment constraint(VAC) as the baseline model. The experimental results show that the screened dataset can achieve better word error rate(WER) than the unscreened dataset.
基金funded by the Ongoing Research Funding Program(ORF-2025-890)King Saud University,Riyadh,Saudi Arabia and was supported by the Competitive Research Fund of theUniversity of Aizu,Japan.
文摘The exponential expansion of the Internet of Things(IoT),Industrial Internet of Things(IIoT),and Transportation Management of Things(TMoT)produces vast amounts of real-time streaming data.Ensuring system dependability,operational efficiency,and security depends on the identification of anomalies in these dynamic and resource-constrained systems.Due to their high computational requirements and inability to efficiently process continuous data streams,traditional anomaly detection techniques often fail in IoT systems.This work presents a resource-efficient adaptive anomaly detection model for real-time streaming data in IoT systems.Extensive experiments were carried out on multiple real-world datasets,achieving an average accuracy score of 96.06%with an execution time close to 7.5 milliseconds for each individual streaming data point,demonstrating its potential for real-time,resourceconstrained applications.The model uses Principal Component Analysis(PCA)for dimensionality reduction and a Z-score technique for anomaly detection.It maintains a low computational footprint with a sliding window mechanism,enabling incremental data processing and identification of both transient and sustained anomalies without storing historical data.The system uses a Multivariate Linear Regression(MLR)based imputation technique that estimates missing or corrupted sensor values,preserving data integrity prior to anomaly detection.The suggested solution is appropriate for many uses in smart cities,industrial automation,environmental monitoring,IoT security,and intelligent transportation systems,and is particularly well-suited for resource-constrained edge devices.
文摘While feeding her ostriches,Wang Xue swiftly responds to fans'questions with a sense of humor through her popular live‑streaming channel.Since Wang,30,an expert in ostrich farming,started live broadcasting on a short‑video platform with her husband Han Peng,their ostrich farming business has flourished,garnering nearly 3 million fans on the platform.“The highest viewership for one single live broadcast reached over 6 million,”said Wang,adding that riding an ostrich at high speed was an essential part of her broadcasts.Through her videos,Wang brought fans close to the ostriches,providing details of their habits,farming methods and economic value.
基金funded by the Joint Project of Industry-University-Research of Jiangsu Province(Grant:BY20231146).
文摘With the widespread application of Internet of Things(IoT)technology,the processing of massive realtime streaming data poses significant challenges to the computational and data-processing capabilities of systems.Although distributed streaming data processing frameworks such asApache Flink andApache Spark Streaming provide solutions,meeting stringent response time requirements while ensuring high throughput and resource utilization remains an urgent problem.To address this,the study proposes a formal modeling approach based on Performance Evaluation Process Algebra(PEPA),which abstracts the core components and interactions of cloud-based distributed streaming data processing systems.Additionally,a generic service flow generation algorithmis introduced,enabling the automatic extraction of service flows fromthe PEPAmodel and the computation of key performance metrics,including response time,throughput,and resource utilization.The novelty of this work lies in the integration of PEPA-based formal modeling with the service flow generation algorithm,bridging the gap between formal modeling and practical performance evaluation for IoT systems.Simulation experiments demonstrate that optimizing the execution efficiency of components can significantly improve system performance.For instance,increasing the task execution rate from 10 to 100 improves system performance by 9.53%,while further increasing it to 200 results in a 21.58%improvement.However,diminishing returns are observed when the execution rate reaches 500,with only a 0.42%gain.Similarly,increasing the number of TaskManagers from 10 to 20 improves response time by 18.49%,but the improvement slows to 6.06% when increasing from 20 to 50,highlighting the importance of co-optimizing component efficiency and resource management to achieve substantial performance gains.This study provides a systematic framework for analyzing and optimizing the performance of IoT systems for large-scale real-time streaming data processing.The proposed approach not only identifies performance bottlenecks but also offers insights into improving system efficiency under different configurations and workloads.
基金Nourah bint Abdulrahman University,Riyadh,Saudi Arabia.supported by the National Research Foundation of Korea(NRF)grant funded by the Korea government(*MSIT)(No.2018R1A5A7059549)the Competitive Research Fund of The University of Aizu,Japan.
文摘Real-time surveillance is attributed to recognizing the variety of actions performed by humans.Human Action Recognition(HAR)is a technique that recognizes human actions from a video stream.A range of variations in human actions makes it difficult to recognize with considerable accuracy.This paper presents a novel deep neural network architecture called Attention RB-Net for HAR using video frames.The input is provided to the model in the form of video frames.The proposed deep architecture is based on the unique structuring of residual blocks with several filter sizes.Features are extracted from each frame via several operations with specific parameters defined in the presented novel Attention-based Residual Bottleneck(Attention-RB)DCNN architecture.A fully connected layer receives an attention-based features matrix,and final classification is performed.Several hyperparameters of the proposed model are initialized using Bayesian Optimization(BO)and later utilized in the trained model for testing.In testing,features are extracted from the self-attention layer and passed to neural network classifiers for the final action classification.Two highly cited datasets,HMDB51 and UCF101,were used to validate the proposed architecture and obtained an average accuracy of 87.70%and 97.30%,respectively.The deep convolutional neural network(DCNN)architecture is compared with state-of-the-art(SOTA)methods,including pre-trained models,inside blocks,and recently published techniques,and performs better.
基金the financial support received from the National Natural Science Foundation of China(Grant No.62174119)the 111 Project (Grant No.B07014)the Foundation for Talent Scientists of Nanchang Institute for Microtechnology of Tianjin University
文摘No-wash bioassays based on nanoparticles are used widely in biochemical procedures because of their responsive sensing and no need forwashing processes.Essential for no-wash biosensing are the interactions between nanoparticles and biomolecules,but it is challenging toachieve controlled bioconjugation of molecules on nanomaterials.Reported here is a way to actively improve nanoparticle-based no-washbioassays by enhancing the binding between biomolecules and gold nanoparticles via acoustic streaming generated by a gigahertz piezoelectricnanoelectromechanical resonator.Tunable micro-vortices are generated at the device-liquid interface,thereby accelerating the internalcirculating flow of the solution,bypassing the diffusion limitation,and thus improving the binding between the biomolecules and goldnanoparticles.Combined with fluorescence quenching,an enhanced and ultrafast no-wash biosensing assay is realized for specific proteins.The sensing method presented here is a versatile tool for different types of biomolecule detection with high efficiency and simplicity.
文摘Many real⁃world machine learning applications face the challenge of dealing with changing data over time,known as concept drift,and the issue of data indeterminacy,where all the true labels available are unrealistic.This can lead to a decrease in the accuracy of the prediction models.The aim of this study is to introduce a new approach for detecting drift,which is based on neutrosophic set theory.This approach takes into account uncertainty in the prediction model and is able to handle indeterminate information,considering its impact on the models performance.The proposed method reads data into windows and calculates a set of values based on the concept of neutrosophic membership.These values are then used in the Neutrosophic Support Vector Machine(N⁃SVM).To address the issue of indeterminate true label data,the values issued by N⁃SVM are expressed as entropy and used as input for the ADWIN(Adaptive Windowing)change detector.When a drift is detected,the prediction model is retrained by including only the most recent instances with the original training data set.The proposed method gives promising results in terms of drift detection accuracy compared to the state of existing drift detection methods such as KSWIN,ADWIN,and DWM.
基金substantially supported by the National Natural Science Foundation of China under Grant No.62002263in part by Tianjin Municipal Education Commission Research Program Project under 2022KJ012Tianjin Science and Technology Program Projects:24YDTPJC00630.
文摘With technological advancements,high-speed rail has emerged as a prevalent mode of transportation.During travel,passengers exhibit a growing demand for streaming media services.However,the high-speed mobile networks environment poses challenges,including frequent base station handoffs,which significantly degrade wireless network transmission performance.Improving transmission efficiency in high-speed mobile networks and optimizing spatiotemporal wireless resource allocation to enhance passengers’media experiences are key research priorities.To address these issues,we propose an Adaptive Cross-Layer Optimization Transmission Method with Environment Awareness(ACOTM-EA)tailored for high-speed rail streaming media.Within this framework,we develop a channel quality prediction model utilizing Kalman filtering and an algorithm to identify packet loss causes.Additionally,we introduce a proactive base station handoffstrategy to minimize handoffrelated disruptions and optimize resource distribution across adjacent base stations.Moreover,this study presents a wireless resource allocation approach based on an enhanced genetic algorithm,coupled with an adaptive bitrate selection mechanism,to maximize passenger Quality of Experience(QoE).To evaluate the proposed method,we designed a simulation experiment and compared ACOTM-EA with established algorithms.Results indicate that ACOTM-EA improves throughput by 11%and enhances passengers’media experience by 5%.
基金supported by the Natural Science Foundation of Sichuan Province(No.2024NSFSC1450)the Fundamental Research Funds for the Central Universities(No.SCU2024D012)the Science and Engineering Connotation Development Project of Sichuan University(No.2020SCUNG129).
文摘In the era of big data,the growing number of real-time data streams often contains a lot of sensitive privacy information.Releasing or sharing this data directly without processing will lead to serious privacy information leakage.This poses a great challenge to conventional privacy protection mechanisms(CPPM).The existing data partitioning methods ignore the number of data replications and information exchanges,resulting in complex distance calculations and inefficient indexing for high-dimensional data.Therefore,CPPM often fails to meet the stringent requirements of efficiency and reliability,especially in dynamic spatiotemporal environments.Addressing this concern,we proposed the Principal Component Enhanced Vantage-point tree(PEV-Tree),which is an enhanced data structure based on the idea of dimension reduction,and constructed a Distributed Spatio-Temporal Privacy Preservation Mechanism(DST-PPM)on it.In this work,principal component analysis and the vantage tree are used to establish the PEV-Tree.In addition,we designed three distributed anonymization algorithms for data streams.These algorithms are named CK-AA,CL-DA,and CT-CA,fulfill the anonymization rules of K-Anonymity,L-Diversity,and T-Closeness,respectively,which have different computational complexities and reliabilities.The higher the complexity,the lower the risk of privacy leakage.DST-PPM can reduce the dimension of high-dimensional information while preserving data characteristics and dividing the data space into vantage points based on distance.It effectively enhances the data processing workflow and increases algorithmefficiency.To verify the validity of the method in this paper,we conducted empirical tests of CK-AA,CL-DA,and CT-CA on conventional datasets and the PEV-Tree,respectively.Based on the big data background of the Internet of Vehicles,we conducted experiments using artificial simulated on-board network data.The results demonstrated that the operational efficiency of the CK-AA,CL-DA,and CT-CA is enhanced by 15.12%,24.55%,and 52.74%,respectively,when deployed on the PEV-Tree.Simultaneously,during homogeneity attacks,the probabilities of information leakage were reduced by 2.31%,1.76%,and 0.19%,respectively.Furthermore,these algorithms showcased superior utility(scalability)when executed across PEV-Trees of varying scales in comparison to their performance on conventional data structures.It indicates that DST-PPM offers marked advantages over CPPM in terms of efficiency,reliability,and scalability.
基金supported by the National Natural Science Foundation of China(No.61931011)the Jiangsu Provincial Key Research and Development Program,China(No.BE2021013-4)the Fundamental Research Project in University Characteristic Disciplines,China(No.ILF240071A24)。
文摘Unmanned Aerial Vehicles(UAVs)coupled with deep learning such as Convolutional Neural Networks(CNNs)have been widely applied across numerous domains,including agriculture,smart city monitoring,and fire rescue operations,owing to their malleability and versatility.However,the computation-intensive and latency-sensitive natures of CNNs present a formidable obstacle to their deployment on resource-constrained UAVs.Some early studies have explored a hybrid approach that dynamically switches between lightweight and complex models to balance accuracy and latency.However,they often overlook scenarios involving multiple concurrent CNN streams,where competition for resources between streams can substantially impact latency and overall system performance.In this paper,we first investigate the deployment of both lightweight and complex models for multiple CNN streams in UAV swarm.Specifically,we formulate an optimization problem to minimize the total latency across multiple CNN streams,under the constraints on UAV memory and the accuracy requirement of each stream.To address this problem,we propose an algorithm called Adaptive Model Switching of collaborative inference for MultiCNN streams(AMSM)to identify the inference strategy with a low latency.Simulation results demonstrate that the proposed AMSM algorithm consistently achieves the lowest latency while meeting the accuracy requirements compared to benchmark algorithms.
文摘On May 14th,following the U.S.adjustment of additional tariffs on Chinese goods,American buyers began stockpiling in earnest.Many cross-border e-commerce companies also received a surge of orders.At 7 PM,a bustling Hangzhou-based cross-border e-commerce company was alive with multiple languages echoing through its live-streaming rooms as backend order numbers climbed steadily.
文摘In the contemporary digital landscape,the proliferation of information has led to an increasing diversity of channels through which consumers obtain information,resulting in a gradual transformation of shopping habits.Consumers now frequently rely on external sources to make well-informed purchasing decisions,leading to the emergence of live shopping as a prominent avenue for gathering product information and completing transactions.E-commerce live streaming has experienced rapid growth,leveraging its ability to generate traffic and capture consumer attention.The integration of content and live streaming not only meets users’psychological needs but also facilitates seamless communication between buyers and sellers.From the perspective of content marketing typologies,this paper examines content marketing across three key dimensions:informational content,entertainment content,and emotional content.It further explores the impact of content marketing on consumers’purchase intentions within the context of e-commerce live streaming.
