With the rapid development of modern information technology,the Internet of Things(IoT)has been integrated into various fields such as social life,industrial production,education,and medical care.Through the connectio...With the rapid development of modern information technology,the Internet of Things(IoT)has been integrated into various fields such as social life,industrial production,education,and medical care.Through the connection of various physical devices,sensors,and machines,it realizes information intercommunication and remote control among devices,significantly enhancing the convenience and efficiency of work and life.However,the rapid development of the IoT has also brought serious security problems.IoT devices have limited resources and a complex network environment,making them one of the important targets of network intrusion attacks.Therefore,from the perspective of deep learning,this paper deeply analyzes the characteristics and key points of IoT intrusion detection,summarizes the application advantages of deep learning in IoT intrusion detection,and proposes application strategies of typical deep learning models in IoT intrusion detection so as to improve the security of the IoT architecture and guarantee people’s convenient lives.展开更多
In recent years,the network public opinion reversal governance events have occurred frequently.Over time,the repeated truth of the matter will not only weaken the rational judgment of the public to a certain extent,so...In recent years,the network public opinion reversal governance events have occurred frequently.Over time,the repeated truth of the matter will not only weaken the rational judgment of the public to a certain extent,so that its negative emotions accumulate,but also have a serious impact on the credibility of the media and the government,and may even further intensify social contradictions.Therefore,in the face of such a complex online public opinion space,accurately identifying the truth behind the incident and how to carry out the reversal of online public opinion governance is particularly critical.And blockchain technology,with its advantages of decentralization and immutable information,provides new technical support for the network public opinion reversal governance.Based on this,this paper gives an overview and analysis of blockchain technology and network public opinion reversal,and on this basis introduces the network public opinion reversal governance mechanism based on blockchain technology,aiming to further optimize the network public opinion reversal governance process,for reference only.展开更多
The Internet of Things technology provides a comprehensive solution for the real-time monitoring of cold chain logistics by integrating sensors,wireless communication,cloud computing,and big data analysis.Based on thi...The Internet of Things technology provides a comprehensive solution for the real-time monitoring of cold chain logistics by integrating sensors,wireless communication,cloud computing,and big data analysis.Based on this,this paper deeply explores the overview and characteristics of the Internet of Things technology,the feasibility analysis of the Internet of Things technology in the cold chain logistics monitoring,the application analysis of the Internet of Things technology in the cold chain logistics real-time monitoring to better improve the management level and operational efficiency of the cold chain logistics,to provide consumers with safer and fresh products.展开更多
Objective:To study the effect of transcranial magnetic stimulation(TMS)on improving motor symptoms in patients with Parkinson’s disease(PD).Methods:60 PD patients who visited the hospital from September 2023 to Augus...Objective:To study the effect of transcranial magnetic stimulation(TMS)on improving motor symptoms in patients with Parkinson’s disease(PD).Methods:60 PD patients who visited the hospital from September 2023 to August 2024 were selected as samples and randomly divided into two groups.Group A received conventional medication plus TMS treatment,while Group B received medication only.The efficacy of motor function improvement,neurological symptoms,mental state,sleep quality,quality of life,and adverse reactions was compared between the two groups.Results:The efficacy of Group A was higher than that of Group B(P<0.05).The scores of the Scales for Outcomes in Parkinson’s Disease-Autonomic(SCOPA-AUT),Mini-Mental State Examination(MMSE),and Pittsburgh Sleep Quality Index(PSQI)in Group A were lower than those in Group B(P<0.05).The quality of life scale(SF-36)score in Group A was higher than that in Group B(P<0.05).The adverse reaction rate in Group A was lower than that in Group B(P<0.05).Conclusion:TMS used in the treatment of PD patients can improve patients’mental state and motor function,optimize sleep quality and quality of life,and is safe and efficient.展开更多
The 2024 MRE HP Special Volume selects papers on new theoretical and experimental developments in the use of static largevolume presses(LVPs)1–3 and dynamic compression4,5 for studies under extreme high-pressure and ...The 2024 MRE HP Special Volume selects papers on new theoretical and experimental developments in the use of static largevolume presses(LVPs)1–3 and dynamic compression4,5 for studies under extreme high-pressure and high-temperature(HPHT)conditions.It also continues the previous year’s6 contemporary focus on superhydrides7–11 with extremely high superconducting temperatures Tc and addresses some controversial issues.12–14 In addition,it explores unconventional pressure-induced chemistry,particularly novel chemical stoichiometry and its impact on geochemistry and cosmochemistry in the deep interiors of Earth and other planets.18–21.展开更多
Ocean geoscience is a highly integrated and interdisciplinary field that plays a critical role in understanding the interaction between Earth’s lithosphere,hydrosphere,atmosphere,biosphere,and anthroposphere.Recent y...Ocean geoscience is a highly integrated and interdisciplinary field that plays a critical role in understanding the interaction between Earth’s lithosphere,hydrosphere,atmosphere,biosphere,and anthroposphere.Recent years have seen tremendous progress in global ocean research,driven by rapid advancements in deep-sea manned and unmanned submersibles,ocean drilling,seafloor observatories,big data assimilation,and supercomputing simulations.Representative examples of breakthroughs are highlighted in this work:(1)Probing sub-seafloor processes.A 10,000-meter ocean-bottom seismometer array has achieved high-resolution imaging of the deepest ocean on the Earth-the Challenger Deep of the Mariana Trench,revealing the role of key tectonic and hydrological processes within the subduction zone.The first sub-ice seafloor seismic and magnetotelluric experiments were successfully conducted at the Arctic Gakkel Ridge,providing significant insights into the dynamics of ultraslow seafloor spreading.(2)Exploration of seafloor resources.Near-seafloor investigations employing underwater robotics and multi-sensor systems have been carried out in areas of hydrothermal vents and cold seeps at global locations,including the Southwest Indian Ridge.These efforts have combined geophysical,oceanographic,chemical,and biological observations with extensive seafloor sampling.(3)Interdisciplinary research of complex catastrophic events.High-resolution simulations integrating ocean observations with supercomputing modeling have made it possible to fully model earthquake-induced seafloor deformation,tsunami propagation,and ocean basin-scale transport of the Fukushima Power Plant-derived radionuclides associated with the 2011 Tohoku earthquake.Among the world’s three major oceans,the Indian Ocean is still relatively underexplored.Major scientific challenges include elucidating crust-mantle interaction,air-sea dynamic coupling,large-scale marine hazards,and responses of ecosystems to major environmental changes,all of which require interdisciplinary collaboration.Future efforts should focus on developing intelligent unmanned observation platform systems,big data and digital twins,and AI-driven hazard modeling.Meanwhile,higher educational reforms should emphasize fostering a new generation of students and young scientists with a solid background and strong critical analysis skills to accelerate technological innovation.展开更多
Accurate and efficient detection of building changes in remote sensing imagery is crucial for urban planning,disaster emergency response,and resource management.However,existing methods face challenges such as spectra...Accurate and efficient detection of building changes in remote sensing imagery is crucial for urban planning,disaster emergency response,and resource management.However,existing methods face challenges such as spectral similarity between buildings and backgrounds,sensor variations,and insufficient computational efficiency.To address these challenges,this paper proposes a novel Multi-scale Efficient Wavelet-based Change Detection Network(MewCDNet),which integrates the advantages of Convolutional Neural Networks and Transformers,balances computational costs,and achieves high-performance building change detection.The network employs EfficientNet-B4 as the backbone for hierarchical feature extraction,integrates multi-level feature maps through a multi-scale fusion strategy,and incorporates two key modules:Cross-temporal Difference Detection(CTDD)and Cross-scale Wavelet Refinement(CSWR).CTDD adopts a dual-branch architecture that combines pixel-wise differencing with semanticaware Euclidean distance weighting to enhance the distinction between true changes and background noise.CSWR integrates Haar-based Discrete Wavelet Transform with multi-head cross-attention mechanisms,enabling cross-scale feature fusion while significantly improving edge localization and suppressing spurious changes.Extensive experiments on four benchmark datasets demonstrate MewCDNet’s superiority over comparison methods:achieving F1 scores of 91.54%on LEVIR,93.70%on WHUCD,and 64.96%on S2Looking for building change detection.Furthermore,MewCDNet exhibits optimal performance on the multi-class⋅SYSU dataset(F1:82.71%),highlighting its exceptional generalization capability.展开更多
High-resolution remote sensing images(HRSIs)are now an essential data source for gathering surface information due to advancements in remote sensing data capture technologies.However,their significant scale changes an...High-resolution remote sensing images(HRSIs)are now an essential data source for gathering surface information due to advancements in remote sensing data capture technologies.However,their significant scale changes and wealth of spatial details pose challenges for semantic segmentation.While convolutional neural networks(CNNs)excel at capturing local features,they are limited in modeling long-range dependencies.Conversely,transformers utilize multihead self-attention to integrate global context effectively,but this approach often incurs a high computational cost.This paper proposes a global-local multiscale context network(GLMCNet)to extract both global and local multiscale contextual information from HRSIs.A detail-enhanced filtering module(DEFM)is proposed at the end of the encoder to refine the encoder outputs further,thereby enhancing the key details extracted by the encoder and effectively suppressing redundant information.In addition,a global-local multiscale transformer block(GLMTB)is proposed in the decoding stage to enable the modeling of rich multiscale global and local information.We also design a stair fusion mechanism to transmit deep semantic information from deep to shallow layers progressively.Finally,we propose the semantic awareness enhancement module(SAEM),which further enhances the representation of multiscale semantic features through spatial attention and covariance channel attention.Extensive ablation analyses and comparative experiments were conducted to evaluate the performance of the proposed method.Specifically,our method achieved a mean Intersection over Union(mIoU)of 86.89%on the ISPRS Potsdam dataset and 84.34%on the ISPRS Vaihingen dataset,outperforming existing models such as ABCNet and BANet.展开更多
In recent years,with the rapid advancement of artificial intelligence,object detection algorithms have made significant strides in accuracy and computational efficiency.Notably,research and applications of Anchor-Free...In recent years,with the rapid advancement of artificial intelligence,object detection algorithms have made significant strides in accuracy and computational efficiency.Notably,research and applications of Anchor-Free models have opened new avenues for real-time target detection in optical remote sensing images(ORSIs).However,in the realmof adversarial attacks,developing adversarial techniques tailored to Anchor-Freemodels remains challenging.Adversarial examples generated based on Anchor-Based models often exhibit poor transferability to these new model architectures.Furthermore,the growing diversity of Anchor-Free models poses additional hurdles to achieving robust transferability of adversarial attacks.This study presents an improved cross-conv-block feature fusion You Only Look Once(YOLO)architecture,meticulously engineered to facilitate the extraction ofmore comprehensive semantic features during the backpropagation process.To address the asymmetry between densely distributed objects in ORSIs and the corresponding detector outputs,a novel dense bounding box attack strategy is proposed.This approach leverages dense target bounding boxes loss in the calculation of adversarial loss functions.Furthermore,by integrating translation-invariant(TI)and momentum-iteration(MI)adversarial methodologies,the proposed framework significantly improves the transferability of adversarial attacks.Experimental results demonstrate that our method achieves superior adversarial attack performance,with adversarial transferability rates(ATR)of 67.53%on the NWPU VHR-10 dataset and 90.71%on the HRSC2016 dataset.Compared to ensemble adversarial attack and cascaded adversarial attack approaches,our method generates adversarial examples in an average of 0.64 s,representing an approximately 14.5%improvement in efficiency under equivalent conditions.展开更多
Deep learning has made significant progress in the field of oriented object detection for remote sensing images.However,existing methods still face challenges when dealing with difficult tasks such as multi-scale targ...Deep learning has made significant progress in the field of oriented object detection for remote sensing images.However,existing methods still face challenges when dealing with difficult tasks such as multi-scale targets,complex backgrounds,and small objects in remote sensing.Maintaining model lightweight to address resource constraints in remote sensing scenarios while improving task completion for remote sensing tasks remains a research hotspot.Therefore,we propose an enhanced multi-scale feature extraction lightweight network EM-YOLO based on the YOLOv8s architecture,specifically optimized for the characteristics of large target scale variations,diverse orientations,and numerous small objects in remote sensing images.Our innovations lie in two main aspects:First,a dynamic snake convolution(DSC)is introduced into the backbone network to enhance the model’s feature extraction capability for oriented targets.Second,an innovative focusing-diffusion module is designed in the feature fusion neck to effectively integrate multi-scale feature information.Finally,we introduce Layer-Adaptive Sparsity for magnitude-based Pruning(LASP)method to perform lightweight network pruning to better complete tasks in resource-constrained scenarios.Experimental results on the lightweight platform Orin demonstrate that the proposed method significantly outperforms the original YOLOv8s model in oriented remote sensing object detection tasks,and achieves comparable or superior performance to state-of-the-art methods on three authoritative remote sensing datasets(DOTA v1.0,DOTA v1.5,and HRSC2016).展开更多
With the accelerating aging process of China’s population,the demand for community elderly care services has shown diversified and personalized characteristics.However,problems such as insufficient total care service...With the accelerating aging process of China’s population,the demand for community elderly care services has shown diversified and personalized characteristics.However,problems such as insufficient total care service resources,uneven distribution,and prominent supply-demand contradictions have seriously affected service quality.Big data technology,with core advantages including data collection,analysis and mining,and accurate prediction,provides a new solution for the allocation of community elderly care service resources.This paper systematically studies the application value of big data technology in the allocation of community elderly care service resources from three aspects:resource allocation efficiency,service accuracy,and management intelligence.Combined with practical needs,it proposes optimal allocation strategies such as building a big data analysis platform and accurately grasping the elderly’s care needs,striving to provide operable path references for the construction of community elderly care service systems,promoting the early realization of the elderly care service goal of“adequate support and proper care for the elderly”,and boosting the high-quality development of China’s elderly care service industry.展开更多
Addressing the widespread issues of internal fragmentation within protected areas and the neglect of surrounding critical habitat networks,this study aims to develop an assessment framework for the precise identificat...Addressing the widespread issues of internal fragmentation within protected areas and the neglect of surrounding critical habitat networks,this study aims to develop an assessment framework for the precise identification and remediation of regional conservation gaps.To this end,we introduce the Framework for Conservation Priority Identification(FCPI).The framework integrates Morphological Spatial Pattern Analysis(MSPA),the Remote Sensing Ecological Index(RSEI),Circuit Theory,and the Minimum Cumulative Resistance(MCR)model to formulate a multidimensional conservation priority index.This index facilitates the identification of critical ecological network components and enables the dynamic prioritization of conservation efforts.A case study of Fuzhou City from 2014 to 2020 reveals that despite an overall improvement in regional environmental quality,the functionality of core ecological sources has markedly declined.Between 2014 and 2020,the number of ecological sources grew by 76.9%,yet their total area shrank by 13.9%.Concurrently,the number of ecological corridors rose from 27 to 53,extending their total length by 380.23 km,which indicates an intensifying trend of habitat fragmentation.Furthermore,a significant number of crucial ecological network nodes,particularly within Minhou County,lie explicitly outside the existing protected area system.This confirms the presence of conservation gaps and unveils the spatiotemporal dynamics of shifting conservation priorities.The research validates that the proposed FCPI can effectively diagnose the dynamic deficiencies within conservation systems.It offers scientific decisionsupport for local governments,facilitating a transition from isolated conservation efforts towards systematic and comprehensive ecological network governance.展开更多
基金the research result of the 2022 Municipal Education Commission Science and Technology Research Plan Project“Research on the Technology of Detecting Double-Surface Cracks in Concrete Lining of Highway Tunnels Based on Image Blast”(KJQN02202403)the first batch of school-level classroom teaching reform projects“Principles Applications of Embedded Systems”(23JG2166)the school-level reform research project“Continuous Results-Oriented Practice Research Based on BOPPPS Teaching Model-Taking the‘Programming Fundamentals’Course as an Example”(22JG332).
文摘With the rapid development of modern information technology,the Internet of Things(IoT)has been integrated into various fields such as social life,industrial production,education,and medical care.Through the connection of various physical devices,sensors,and machines,it realizes information intercommunication and remote control among devices,significantly enhancing the convenience and efficiency of work and life.However,the rapid development of the IoT has also brought serious security problems.IoT devices have limited resources and a complex network environment,making them one of the important targets of network intrusion attacks.Therefore,from the perspective of deep learning,this paper deeply analyzes the characteristics and key points of IoT intrusion detection,summarizes the application advantages of deep learning in IoT intrusion detection,and proposes application strategies of typical deep learning models in IoT intrusion detection so as to improve the security of the IoT architecture and guarantee people’s convenient lives.
基金Gansu Provincial Department of Education:Innovation Fund Project for College Teachers:Research on the Governance of Network Public Opinion Reversal Based on Blockchain Technology(NO.2025A-137)Teaching Reform Research at the School Level of Gansu University of Political Science and Law in 2024:Research and Implementation of CSCW Cloud Storage Collective Lesson Preparation System Based on Blockchain Technology(NO.GZJG2024-A04)+1 种基金2024 Gansu University of Political Science and Law“Three in One Education”Research:Project Research and Implementation of CSCW Network Teaching Platform Based on Web in the Direction of Network Education(NO.GZSQYR-37)2024 Lanzhou Philosophy and Social Sciences Planning Project:Research on Cloud Governance of Network Public Opinion for Major Public Events in Lanzhou City(NO.24-B45)。
文摘In recent years,the network public opinion reversal governance events have occurred frequently.Over time,the repeated truth of the matter will not only weaken the rational judgment of the public to a certain extent,so that its negative emotions accumulate,but also have a serious impact on the credibility of the media and the government,and may even further intensify social contradictions.Therefore,in the face of such a complex online public opinion space,accurately identifying the truth behind the incident and how to carry out the reversal of online public opinion governance is particularly critical.And blockchain technology,with its advantages of decentralization and immutable information,provides new technical support for the network public opinion reversal governance.Based on this,this paper gives an overview and analysis of blockchain technology and network public opinion reversal,and on this basis introduces the network public opinion reversal governance mechanism based on blockchain technology,aiming to further optimize the network public opinion reversal governance process,for reference only.
文摘The Internet of Things technology provides a comprehensive solution for the real-time monitoring of cold chain logistics by integrating sensors,wireless communication,cloud computing,and big data analysis.Based on this,this paper deeply explores the overview and characteristics of the Internet of Things technology,the feasibility analysis of the Internet of Things technology in the cold chain logistics monitoring,the application analysis of the Internet of Things technology in the cold chain logistics real-time monitoring to better improve the management level and operational efficiency of the cold chain logistics,to provide consumers with safer and fresh products.
文摘Objective:To study the effect of transcranial magnetic stimulation(TMS)on improving motor symptoms in patients with Parkinson’s disease(PD).Methods:60 PD patients who visited the hospital from September 2023 to August 2024 were selected as samples and randomly divided into two groups.Group A received conventional medication plus TMS treatment,while Group B received medication only.The efficacy of motor function improvement,neurological symptoms,mental state,sleep quality,quality of life,and adverse reactions was compared between the two groups.Results:The efficacy of Group A was higher than that of Group B(P<0.05).The scores of the Scales for Outcomes in Parkinson’s Disease-Autonomic(SCOPA-AUT),Mini-Mental State Examination(MMSE),and Pittsburgh Sleep Quality Index(PSQI)in Group A were lower than those in Group B(P<0.05).The quality of life scale(SF-36)score in Group A was higher than that in Group B(P<0.05).The adverse reaction rate in Group A was lower than that in Group B(P<0.05).Conclusion:TMS used in the treatment of PD patients can improve patients’mental state and motor function,optimize sleep quality and quality of life,and is safe and efficient.
基金financial support from the Shanghai Key Laboratory of MFree,China(Grant No.22dz2260800)the Shanghai Science and Technology Committee,China(Grant No.22JC1410300).
文摘The 2024 MRE HP Special Volume selects papers on new theoretical and experimental developments in the use of static largevolume presses(LVPs)1–3 and dynamic compression4,5 for studies under extreme high-pressure and high-temperature(HPHT)conditions.It also continues the previous year’s6 contemporary focus on superhydrides7–11 with extremely high superconducting temperatures Tc and addresses some controversial issues.12–14 In addition,it explores unconventional pressure-induced chemistry,particularly novel chemical stoichiometry and its impact on geochemistry and cosmochemistry in the deep interiors of Earth and other planets.18–21.
基金supported by the National Natural Science Foundation of China(Grant No.92258303)the National Key Research and Development Program of China(Grant Nos.2024YFF0506704 and 2023YFF0803404).
文摘Ocean geoscience is a highly integrated and interdisciplinary field that plays a critical role in understanding the interaction between Earth’s lithosphere,hydrosphere,atmosphere,biosphere,and anthroposphere.Recent years have seen tremendous progress in global ocean research,driven by rapid advancements in deep-sea manned and unmanned submersibles,ocean drilling,seafloor observatories,big data assimilation,and supercomputing simulations.Representative examples of breakthroughs are highlighted in this work:(1)Probing sub-seafloor processes.A 10,000-meter ocean-bottom seismometer array has achieved high-resolution imaging of the deepest ocean on the Earth-the Challenger Deep of the Mariana Trench,revealing the role of key tectonic and hydrological processes within the subduction zone.The first sub-ice seafloor seismic and magnetotelluric experiments were successfully conducted at the Arctic Gakkel Ridge,providing significant insights into the dynamics of ultraslow seafloor spreading.(2)Exploration of seafloor resources.Near-seafloor investigations employing underwater robotics and multi-sensor systems have been carried out in areas of hydrothermal vents and cold seeps at global locations,including the Southwest Indian Ridge.These efforts have combined geophysical,oceanographic,chemical,and biological observations with extensive seafloor sampling.(3)Interdisciplinary research of complex catastrophic events.High-resolution simulations integrating ocean observations with supercomputing modeling have made it possible to fully model earthquake-induced seafloor deformation,tsunami propagation,and ocean basin-scale transport of the Fukushima Power Plant-derived radionuclides associated with the 2011 Tohoku earthquake.Among the world’s three major oceans,the Indian Ocean is still relatively underexplored.Major scientific challenges include elucidating crust-mantle interaction,air-sea dynamic coupling,large-scale marine hazards,and responses of ecosystems to major environmental changes,all of which require interdisciplinary collaboration.Future efforts should focus on developing intelligent unmanned observation platform systems,big data and digital twins,and AI-driven hazard modeling.Meanwhile,higher educational reforms should emphasize fostering a new generation of students and young scientists with a solid background and strong critical analysis skills to accelerate technological innovation.
基金supported by the Henan Province Key R&D Project under Grant 241111210400the Henan Provincial Science and Technology Research Project under Grants 252102211047,252102211062,252102211055 and 232102210069+2 种基金the Jiangsu Provincial Scheme Double Initiative Plan JSS-CBS20230474,the XJTLU RDF-21-02-008the Science and Technology Innovation Project of Zhengzhou University of Light Industry under Grant 23XNKJTD0205the Higher Education Teaching Reform Research and Practice Project of Henan Province under Grant 2024SJGLX0126。
文摘Accurate and efficient detection of building changes in remote sensing imagery is crucial for urban planning,disaster emergency response,and resource management.However,existing methods face challenges such as spectral similarity between buildings and backgrounds,sensor variations,and insufficient computational efficiency.To address these challenges,this paper proposes a novel Multi-scale Efficient Wavelet-based Change Detection Network(MewCDNet),which integrates the advantages of Convolutional Neural Networks and Transformers,balances computational costs,and achieves high-performance building change detection.The network employs EfficientNet-B4 as the backbone for hierarchical feature extraction,integrates multi-level feature maps through a multi-scale fusion strategy,and incorporates two key modules:Cross-temporal Difference Detection(CTDD)and Cross-scale Wavelet Refinement(CSWR).CTDD adopts a dual-branch architecture that combines pixel-wise differencing with semanticaware Euclidean distance weighting to enhance the distinction between true changes and background noise.CSWR integrates Haar-based Discrete Wavelet Transform with multi-head cross-attention mechanisms,enabling cross-scale feature fusion while significantly improving edge localization and suppressing spurious changes.Extensive experiments on four benchmark datasets demonstrate MewCDNet’s superiority over comparison methods:achieving F1 scores of 91.54%on LEVIR,93.70%on WHUCD,and 64.96%on S2Looking for building change detection.Furthermore,MewCDNet exhibits optimal performance on the multi-class⋅SYSU dataset(F1:82.71%),highlighting its exceptional generalization capability.
基金provided by the Science Research Project of Hebei Education Department under grant No.BJK2024115.
文摘High-resolution remote sensing images(HRSIs)are now an essential data source for gathering surface information due to advancements in remote sensing data capture technologies.However,their significant scale changes and wealth of spatial details pose challenges for semantic segmentation.While convolutional neural networks(CNNs)excel at capturing local features,they are limited in modeling long-range dependencies.Conversely,transformers utilize multihead self-attention to integrate global context effectively,but this approach often incurs a high computational cost.This paper proposes a global-local multiscale context network(GLMCNet)to extract both global and local multiscale contextual information from HRSIs.A detail-enhanced filtering module(DEFM)is proposed at the end of the encoder to refine the encoder outputs further,thereby enhancing the key details extracted by the encoder and effectively suppressing redundant information.In addition,a global-local multiscale transformer block(GLMTB)is proposed in the decoding stage to enable the modeling of rich multiscale global and local information.We also design a stair fusion mechanism to transmit deep semantic information from deep to shallow layers progressively.Finally,we propose the semantic awareness enhancement module(SAEM),which further enhances the representation of multiscale semantic features through spatial attention and covariance channel attention.Extensive ablation analyses and comparative experiments were conducted to evaluate the performance of the proposed method.Specifically,our method achieved a mean Intersection over Union(mIoU)of 86.89%on the ISPRS Potsdam dataset and 84.34%on the ISPRS Vaihingen dataset,outperforming existing models such as ABCNet and BANet.
文摘In recent years,with the rapid advancement of artificial intelligence,object detection algorithms have made significant strides in accuracy and computational efficiency.Notably,research and applications of Anchor-Free models have opened new avenues for real-time target detection in optical remote sensing images(ORSIs).However,in the realmof adversarial attacks,developing adversarial techniques tailored to Anchor-Freemodels remains challenging.Adversarial examples generated based on Anchor-Based models often exhibit poor transferability to these new model architectures.Furthermore,the growing diversity of Anchor-Free models poses additional hurdles to achieving robust transferability of adversarial attacks.This study presents an improved cross-conv-block feature fusion You Only Look Once(YOLO)architecture,meticulously engineered to facilitate the extraction ofmore comprehensive semantic features during the backpropagation process.To address the asymmetry between densely distributed objects in ORSIs and the corresponding detector outputs,a novel dense bounding box attack strategy is proposed.This approach leverages dense target bounding boxes loss in the calculation of adversarial loss functions.Furthermore,by integrating translation-invariant(TI)and momentum-iteration(MI)adversarial methodologies,the proposed framework significantly improves the transferability of adversarial attacks.Experimental results demonstrate that our method achieves superior adversarial attack performance,with adversarial transferability rates(ATR)of 67.53%on the NWPU VHR-10 dataset and 90.71%on the HRSC2016 dataset.Compared to ensemble adversarial attack and cascaded adversarial attack approaches,our method generates adversarial examples in an average of 0.64 s,representing an approximately 14.5%improvement in efficiency under equivalent conditions.
基金funded by the Hainan Province Science and Technology Special Fund under Grant ZDYF2024GXJS292.
文摘Deep learning has made significant progress in the field of oriented object detection for remote sensing images.However,existing methods still face challenges when dealing with difficult tasks such as multi-scale targets,complex backgrounds,and small objects in remote sensing.Maintaining model lightweight to address resource constraints in remote sensing scenarios while improving task completion for remote sensing tasks remains a research hotspot.Therefore,we propose an enhanced multi-scale feature extraction lightweight network EM-YOLO based on the YOLOv8s architecture,specifically optimized for the characteristics of large target scale variations,diverse orientations,and numerous small objects in remote sensing images.Our innovations lie in two main aspects:First,a dynamic snake convolution(DSC)is introduced into the backbone network to enhance the model’s feature extraction capability for oriented targets.Second,an innovative focusing-diffusion module is designed in the feature fusion neck to effectively integrate multi-scale feature information.Finally,we introduce Layer-Adaptive Sparsity for magnitude-based Pruning(LASP)method to perform lightweight network pruning to better complete tasks in resource-constrained scenarios.Experimental results on the lightweight platform Orin demonstrate that the proposed method significantly outperforms the original YOLOv8s model in oriented remote sensing object detection tasks,and achieves comparable or superior performance to state-of-the-art methods on three authoritative remote sensing datasets(DOTA v1.0,DOTA v1.5,and HRSC2016).
文摘With the accelerating aging process of China’s population,the demand for community elderly care services has shown diversified and personalized characteristics.However,problems such as insufficient total care service resources,uneven distribution,and prominent supply-demand contradictions have seriously affected service quality.Big data technology,with core advantages including data collection,analysis and mining,and accurate prediction,provides a new solution for the allocation of community elderly care service resources.This paper systematically studies the application value of big data technology in the allocation of community elderly care service resources from three aspects:resource allocation efficiency,service accuracy,and management intelligence.Combined with practical needs,it proposes optimal allocation strategies such as building a big data analysis platform and accurately grasping the elderly’s care needs,striving to provide operable path references for the construction of community elderly care service systems,promoting the early realization of the elderly care service goal of“adequate support and proper care for the elderly”,and boosting the high-quality development of China’s elderly care service industry.
基金supported by the Natural Science Foundation of Fujian Province(2023J01434)the Science and Technology Innovation Special Fund Project of Fujian Agriculture and Forestry University(KFb22028XA)。
文摘Addressing the widespread issues of internal fragmentation within protected areas and the neglect of surrounding critical habitat networks,this study aims to develop an assessment framework for the precise identification and remediation of regional conservation gaps.To this end,we introduce the Framework for Conservation Priority Identification(FCPI).The framework integrates Morphological Spatial Pattern Analysis(MSPA),the Remote Sensing Ecological Index(RSEI),Circuit Theory,and the Minimum Cumulative Resistance(MCR)model to formulate a multidimensional conservation priority index.This index facilitates the identification of critical ecological network components and enables the dynamic prioritization of conservation efforts.A case study of Fuzhou City from 2014 to 2020 reveals that despite an overall improvement in regional environmental quality,the functionality of core ecological sources has markedly declined.Between 2014 and 2020,the number of ecological sources grew by 76.9%,yet their total area shrank by 13.9%.Concurrently,the number of ecological corridors rose from 27 to 53,extending their total length by 380.23 km,which indicates an intensifying trend of habitat fragmentation.Furthermore,a significant number of crucial ecological network nodes,particularly within Minhou County,lie explicitly outside the existing protected area system.This confirms the presence of conservation gaps and unveils the spatiotemporal dynamics of shifting conservation priorities.The research validates that the proposed FCPI can effectively diagnose the dynamic deficiencies within conservation systems.It offers scientific decisionsupport for local governments,facilitating a transition from isolated conservation efforts towards systematic and comprehensive ecological network governance.
