检索结果-南通市图书馆

CAAI Transactions on Intelligence Technology 2024年第1期9卷 239-249页

作者： Shuo Zhao peng cui Jing Shen Haibo Liu School of Computer Science and Technology Harbin University of Science and TechnologyHarbinChina School of Computer Science and Technology Harbin Engineering UniversityHarbinChina

Recently,weak supervision has received growing attention in the field of salient object detection due to the convenience of labelling.However,there is a large performance gap between weakly supervised and fully supervised salient object detectors because the scribble annotation can only provide very limited foreground/background information.Therefore,an intuitive idea is to infer annotations that cover more complete object and background regions for training.To this end,a label inference strategy is proposed based on the assumption that pixels with similar colours and close positions should have consistent labels.Specifically,k-means clustering algorithm was first performed on both colours and coordinates of original annotations,and then assigned the same labels to points having similar colours with colour cluster centres and near coordinate cluster centres.Next,the same annotations for pixels with similar colours within each kernel neighbourhood was set further.Extensive experiments on six benchmarks demonstrate that our method can significantly improve the performance and achieve the state-of-the-art results.

关键词： label inference salient object detection weak supervision

A new scheme of climate-vegetation regionalization in the Hengduan Mountains Region

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science China Earth Sciences 2024年第3期67卷 751-768页

作者： Xueqin ZHANG Xiaoming XU Xiang LI peng cui Du ZHENG Key Laboratory of Land Surface Pattern and Simulation Institute of Geographic Sciences and Natural Resources ResearchChinese Academy of SciencesBeijing 100101China University of Chinese Academy of Sciences Beijing 100049China China-Pakistan Joint Research Center on Earth Sciences Chinese Academy of Sciences-Higher Education Commission of PakistanIslamabad 45320Pakistan

The Hengduan Mountains Region(HMR) is essential for the future ecological protection, clean energy production,Sichuan-Xizang and Yunnan-Xizang railways, and other major infrastructure projects in China. The distributions of climate and vegetation exhibit significant regional differentiation and vertical zonality due to the rugged longitudinal ranges and gorges and the complex disaster-prone environments in HMR. Therefore, it is urgent to develop the climate-vegetation regionalization in HMR to effectively satisfy the national requirements such as agricultural production and ecological protection, mountain disaster risk prevention, and major project construction. We here develop a new scheme of climate-vegetation regionalization with the latest demarcation outcome of HMR, the ground observation from 122 meteorological stations in HMR and its surrounding areas during 1990–2019, and the high-precision remote sensing data of land cover types. The new scheme first constructs the regionalization index system, fully considering the extraordinarily complicated geomorphic pattern of mountains and valleys, the scarcity of meteorological observations, and the remarkable differentiation of climate and vegetation in HMR. The system consists of three primary regionalization indices(i.e., days with daily average temperature steady above 10°C, aridity index, and main vegetation types, dividing the temperature zones, moisture regions, and vegetation subregions, respectively) and three auxiliary indices of the accumulated temperature above 10°C, and the temperatures in January and July. Then, the HMR is divided into five temperature zones, 20 moisture regions, and 55 vegetation subregions. Compared with previous regionalization schemes, the new scheme optimizes the climate spatial interpolation model of thin plate smoothing spline suitable for the unique terrain in HMR. Moreover, the disputed division index threshold between different climatic zones(regions) is scientifically clarified using geographical detectors. Specifically, t

关键词： Climate-vegetation regionalization Thin plate smoothing spline(TPSS) GeoDetector Stepwise downscaling pane division method Hengduan Mountains Region

Quick and automatic detection of co-seismic landslides with multifeature deep learning model

在线全文

学校读者我要写书评

暂无评论

Science China Earth Sciences 2024年第7期67卷 2311-2325页

作者： Wenchao HUANGFU Haijun QIU peng cui Dongdong YANG Ya LIU Bingzhe TANG Zijing LIU Mohib ULLAH College of Urban and Environmental Sciences Northwest UniversityXi'an 710127China Institute of Mountain Hazards and Environment Chinese Academy of SciencesChengdu 610041China Shaanxi Key Laboratory of Earth Surface System and Environmental Carrying Capacity Northwest UniversityXi'an 710127China Institute of Earth Surface System and Hazards Northwest UniversityXi'an 710127China

Co-seismic landslide detection is essential for post-disaster rescue and risk assessment after an earthquake event.However,a variety of ground objects,including roads and bare land,have spectral characteristics similar to those of co-seismic landslides,making it difficult to gather information and assess their impact rapidly and accurately.Therefore,an automatic detection method based on a deep learning model,named ENVINet5,with multiple features(ENVINet5_MF)was proposed to solve this problem and improve the detection accuracy of co-seismic landslides.The ENVINet5_MF method is advantageous for co-seismic landslide detection because it features a landslide gain index(LGI)that effectively eliminates the spectral interference of bare land and roads.We conducted two experiments using multi-temporal PlanetScope images acquired in Hokkaido,Japan,and Mainling,China.The accuracy evaluation and rationality analysis show that ENVINet5_MF performed better than comparative methods and that the co-seismic landslide areas detected by ENVINet5_MF were the most consistent with ground reference data.The findings of this study suggest that ENVINet5_MF can provide an efficient and accurate method for coseismic landslide detection to ensure a rapid response to co-seismic landslide disasters.

关键词： Co-seismic landslide Automatic detection Deep learning Landslide gain index PlanetScope images

Deep learning for cross-region streamflow and flood forecasting at a global scale

在线全文

学校读者我要写书评

暂无评论

The Innovation 2024年第3期5卷 99-109页

作者： Binlan Zhang Chaojun Ouyang peng cui Qingsong Xu Dongpo Wang Fei Zhang Zhong Li Linfeng Fan Marco Lovati Yanling Liu Qianqian Zhang State Key Laboratory of Mountain Hazards and Engineering Resilience Institute of Mountain Hazards and EnvironmentChinese Academy of SciencesChengdu 610299China State Key Laboratory of Geohazard Prevention and Geoenvironment Protection Chengdu University of TechnologyChengdu 610059China Data Science in Earth Observation Technical University of Munich80333 MunichGermany State Key Laboratory of Loess and Quaternary Geology Institute of Earth EnvironmentChinese Academy of SciencesXi’an 710061China Department of Civil Engineering McMaster UniversityHamiltonON L8S 4L8Canada Chengdu University of Information Technology Chengdu 610225China

Streamflow and flood forecasting remains one of the long-standing challenges in hydrology.Traditional physically based models are hampered by sparse parameters and complex calibration procedures particularly in ungauged catchments.We propose a novel hybrid deep learning model termed encoder-decoder double-layer long short-term memory(ED-DLSTM)to address streamflow forecasting at global scale for all(gauged and ungauged)catchments.Using historical datasets,ED-DLSTM yields a mean Nash-Sutcliffe efficiency coefficient(NSE)of 0.75 across more than 2,000 catchments from the United States,Canada,Central Europe,and the United Kingdom,highlighting improvements by the state-of-the-art machine learning over traditional hydrologic models.Moreover,ED-DLSTM is applied to 160 ungauged catchments in Chile and 76.9%of catchments obtain NSE>0 in the best situation.The interpretability of cross-region capacities of ED-DLSTM are established through the cell state induced by adding a spatial attribute encoding module,which can spontaneously form hydrological regionalization effects after performing spatial coding for different catchments.The study demonstrates the potential of deep leaning methods to overcome the ubiquitous lack of hydrologic information and deficiencies in physical model structure and parameterization.

关键词： forecasting flood flow

A GNSS Anti-Spoofing Technique Based on the Spatial Distribution Characteristic of the Residual Vectors

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Tsinghua Science and Technology 2024年第2期29卷 457-468页

作者： Qi'an Wu Xiaowei cui Mingquan Lu pengxiang Yang peng Wu Department of Electronic Enginnering Tsinghua UniversityBeijing 100084China Xi’an Modern Control Technology Research Institute Xi’an 710065China

Anti-spoofing is becoming a crucial technique for applications with high navigation accuracy and reliability requirements.Anti-spoofing technique based on Receiver Autonomous Integrity Monitoring(RAIM)is a good choice for most Global Navigation Satellite System(GNSS)receivers because it does not require any change to the hardware of the receiver.However,the conventional RAIM method can only detect and mitigate a single spoofing signal.Some improved RAIM methods can deal with more spoofing signals,but the computational complexity increases dramatically when the number of satellites in view increase or need additional information.This paper proposes a new RAIM method,called the SRV-RAIM method,which has a very low computation complexity regardless of the number of satellites in view and can deal with any number of spoofing signals.The key to the new method is the spatial distribution characteristic of the Satellites'Residual Vectors(SRV).In replay or generative spoofing scenarios,the pseudorange measurements of spoofing signals are consistent,the residual vectors of real satellites and those of spoofing satellites have good separation characteristics in spatial distribution.Based on this characteristic,the SRV-RAIM method is proposed,and the simulation results show that the method can separate the real signals and the spoofing signals with an average probability of 86.55%in the case of 12 visible satellites,regardless of the number of spoofing signals.Compared to the conventional traversal-RAIM method,the performance is only reduced by 3.59%,but the computational cost is reduced by 98.3%,so most of the GNSS receivers can run the SRV-RAIM algorithm in time.

关键词： Global Navigation Satellite System(GNSS) anti-spoofing Receiver Autonomous Integrity Monitoring(RAIM) Satellites'Residual Vectors(SRV)

The Combined Effects of High Temperatures and Ozone Pollution on Medical Emergency Calls—Jinan City,Shandong Province,China,2013–2019

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

China CDC weekly 2024年第2期6卷 30-35,I0001-I0007页

作者： Huiyun Chang Chong Liu Xingyi Geng Xiumiao peng Ke Huang Liheng Wang Liangliang cui Department of Scientific Research and Education Jinan Municipal Center for Disease Control and PreventionJinan Municipal Center for Disease Control and Prevention Affiliated to Shandong UniversityJinan CityShandong ProvinceChina Jinan Health Well Development Center Jinan CityShandong ProvinceChina Department of Environmental Health Jinan Municipal Center for Disease Control and PreventionJinan CityShandong ProvinceChina Department of Emergency Jinan Medical Emergency CenterJinan CityShandong ProvinceChina

What is already known about this topic?Studies have extensively documented the separate and independent effects of extreme temperature and ozone on morbidity and mortality associated with respiratory and circulatory diseases.What is added by this report?The study revealed a significant association between elevated temperature,ozone pollution,and the combined effect of high temperature and ozone pollution with an increased risk of all-cause medical emergency calls(MECs)and MECs specifically related to neurological diseases.What are the implications for public health practice?Interventional measures should be implemented to mitigate exposure to high temperatures and ozone levels.Specifically,during the warm season,it is crucial for relevant authorities to focus on disseminating scientific information regarding the health impacts of elevated temperatures and ozone pollution.Additionally,timely public health advisories should be issued to alert the public effectively.

关键词： ozone elevated pollution

Multiwavelength Observations of the Infrared Dust Bubble N75 and its Surroundings

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Research in Astronomy and Astrophysics 2024年第4期24卷 262-270页

作者： Quan-Ling cui Chuan-peng Zhang Jun-Jie Wang National Astronomical Observatories Chinese Academy of SciencesBeijing 100101China University of Chinese Academy of Sciences Beijing 100049China College of Science Qiqihar UniversityQiqihar 161006China Guizhou Radio Astronomical Observatory Guizhou UniversityGuiyang 550000China

Infrared dust bubbles play an important role in the study of star formation and the evolution of the interstellar medium.In this work,we study the infrared dust bubble N75 and the infrared dark cloud G38.93 mainly using the tracers C^(18)O,HCO^(+),HNC and N_(2)H^(+)observed by the 30 m IRAM telescope.We also study the targets using data from large-scale surveys:GLIMPSE,MIPSGAL,GRS,NRAO VLA Sky Survey and Bolocam Galactic Plane Survey.We found that the C^(18)O emission is morphologically similar to the Spitzer IRAC 8.0μm emission.The1.1 mm cold dust emission of G38.93 shows an elongated structure from southwest to northeast.The ionized gas from G38.93 is surrounded by polycyclic aromatic hydrocarbon emission,which may be excited by radiation from G38.93.We found that the identified young stellar objects tend to cluster around G38.93 and are mostly in class II with several class I cases distributed around N75,but no class II examples.We also found evidence of expanding feedback,which could have triggered star formation.

关键词： ISM:bubbles infrared:ISM galaxies:star formation

Possible Habitats for NH_(3),NH_(2)D,H^(13)CN,HC^(15)N,SO,and C^(18)O in the Initial Conditions of High-mass Star Formation

在线全文

学校读者我要写书评

暂无评论

Research in Astronomy and Astrophysics 2024年第4期24卷 237-252页

The initial condition of high-mass star formation is a complex area of study because of the high densities(n_(H_(2))>106cm^(-3))and low temperatures(T_(dust)<18 K)involved.Under such conditions,many molecules become depleted from the gas phase by freezing out onto dust grains.However,the N-bearing and deuterated species could remain gaseous under these extreme conditions,suggesting that they may serve as ideal tracers.In this paper,using the Plateau de Bure Interferometer and Very Large Array observations at 1.3 mm,3.5 mm,and 1.3 cm,we investigate the possible habitats for NH_(3),NH_(2)D,H^(13)CN,HC^(15)N,SO,and C^(18)O in eight massive precluster and protocluster clumps G18.17,G18.21,G23.97N,G23.98,G23.44,G23.97S,G25.38,and G25.71.We found that the NH3cores are in good agreement with the 3.5 mm peak emission,but the NH_(3)is much more extended than the 3.5 mm emission structure.The SO distributions agree well with the 3.5 mm peaks for the evolved star formation stage,but we did not detect any SO emission in the four earliest star formation sources.C^(18)O is a poor tracer in conditions of the cold(■18 K)and dense(■10^(4)cm^(-3))cores,e.g.,the prestellar cores.We also found that the NH_(2)D cores are mainly located in the temperature range of 13.0-20.0 K,and the NH_(2)D lines may be strongly depleted above 20 K.

关键词： galaxies:star formation techniques:interferometric methods:observational

User Preference Aware Hierarchical Edge-User Cooperative Caching Strategy

在线全文

学校读者我要写书评

暂无评论

China Communications 2024年第6期21卷 69-86页

作者： Wu Dapeng Yang Lin cui Yaping He peng Wang Ruyan School of Communications and Information Engineering Chongqing University of Posts and TelecommunicationsChongqing 400065China Advanced Network and Intelligent Connection Technology Key Laboratory of Chongqing Education Commission of China Chongqing 400065China Chongqing Key Laboratory of Ubiquitous Sensing and Networking Chongqing 400065China

The emergence of various new services has posed a huge challenge to the existing network architecture.To improve the network delay and backhaul pressure,caching popular contents at the edge of network has been considered as a feasible scheme.However,how to efficiently utilize the limited caching resources to cache diverse contents has been confirmed as a tough problem in the past decade.In this paper,considering the time-varying user requests and the heterogeneous content sizes,a user preference aware hierarchical cooperative caching strategy in edge-user caching architecture is proposed.We divide the caching strategy into three phases,that is,the content placement,the content delivery and the content update.In the content placement phase,a cooperative content placement algorithm for local content popularity is designed to cache contents proactively.In the content delivery phase,a cooperative delivery algorithm is proposed to deliver the cached contents.In the content update phase,a content update algorithm is proposed according to the popularity of the contents.Finally,the proposed caching strategy is validated using the MovieLens dataset,and the results reveal that the proposed strategy improves the delay performance by at least 35.3%compared with the other three benchmark strategies.

关键词： cooperative caching network delay timevarying popularity user preference

NFT Security Matrix:Towards Modeling NFT Ecosystem Threat

在线全文

学校读者我要写书评

暂无评论

Computer Modeling in Engineering & Sciences 2024年第6期139卷 3255-3285页

作者： peng Liao Chaoge Liu Jie Yin Zhi Wang Xiang cui Key Laboratory of Trustworthy Distributed Computing and Service Beijing University of Posts and TelecommunicationsBeijing100876China Zhongguancun Laboratory Beijing100094China Institute of Information Engineering Chinese Academy of SciencesBeijing100085China

Digital assets have boomed over the past few years with the emergence of Non-fungible Tokens(NFTs).To be specific,the total trading volume of digital assets reached an astounding$55.5 billion in 2022.Nevertheless,numerous security concerns have been raised by the rapid expansion of the NFT ecosystem.NFT holders are exposed to a plethora of scams and traps,putting their digital assets at risk of being lost.However,academic research on NFT security is scarce,and the security issues have aroused rare attention.In this study,the NFT ecological process is comprehensively explored.This process falls into five different stages encompassing the entire lifecycle of NFTs.Subsequently,the security issues regarding the respective stage are elaborated and analyzed in depth.A matrix model is proposed as a novel contribution to the categorization of NFT security issues.Diverse data are collected from social networks,the Ethereum blockchain,and NFT markets to substantiate our claims regarding the severity of security concerns in the NFT ecosystem.From this comprehensive dataset,nine key NFT security issues are identified from the matrix model and then subjected to qualitative and quantitative analysis.This study aims to shed light on the severity of NFT ecosystem security issues.The findings stress the need for increased attention and proactive measures to safeguard the NFT ecosystem.

关键词： Non-fungible token blockchain cyber security