检索结果-内蒙古大学图书馆

GP-PSENet: a group-related dilated and a parallel extensional dilation-wise residual encoder for scene text detection

引用

Neural computing and applications 2025年第10期37卷 7159-7181页

作者： Huang, Liwen Yang, Wenyuan Fujian Key Laboratory of Granular Computing and Application Minnan Normal University Fujian Zhangzhou363000 China School of Mathematics and Statistics Minnan Normal University Fujian Zhangzhou363000 China

At present, scene text detection is grabbing more and more attention as an offshoot of machine vision. However, due to the existing long types of text instances and complex background context, less exact localization and higher missed detection cases still remain in text detection domain. Accordingly, with the aim of tackling these two issues, we propose a text detector named GP-PSENet that comprises a combination of a group-related dilated encoder, a parallel extensional dilation-wise residual encoder and a mixed upsample. Firstly, feature maps of the lowest level processed by the backbone network are sent to a dilated encoder with group linkage. And the group residual module provides stratification to join group coefficients and dilated factors. This module can enhance the correctness of predictions about longer boundary boxes. Secondly, semantic information from the highest level is fed into a parallel extensional dilation-wise residual encoder. The extensional dilation-wise module is capable of obtaining diverse receptive fields by more parallel branches. And it can alleviate error detection from interfering material in the background. Thirdly, the feature maps processed in the second step are given to the mixed upsample module for transforming so as to the next fuse. Finally, the processed two-level feature maps are fused and sent to the progressive scale expansion algorithm for the final post-processing to gain the predicted coordinate points. Ablation experiments are conducted on CTW1500, ICDAR15, MSRA-TD500 and Total-Text datasets to confirm the availability of the proposed method. The values of precision on these datasets reach 86.24%, 87.84%, 73.98% and 90.48%. The proposed method is also competitive with other scene detection methods. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2025.

关键词： Computer vision

来源：评论

学校读者我要写书评

暂无评论

High-order finite volume methods for solving compressible multicomponent flows

引用

Science China Mathematics 2025年第3期68卷 729-760页

作者： Feng Zheng Jianxian Qiu School of Mathematics and Statistics Fujian Key Laboratory of Mathematical Analysis and ApplicationFujian Normal UniversityFuzhou 350117China School of Mathematical Sciences and Fujian Provincial Key Laboratory of Mathematical Modeling&High-Performance Scientific Computing Xiamen UniversityXiamen 361005China

In this paper, we propose a high-order finite volume method for solving multicomponent fluid problems. Our method couples the quasi-conservative form with the reconstruction of conservative variables in a characteristic manner. The source term and numerical fluxes are carefully designed to maintain the pressure and velocity equilibrium for the interface-only problem and preserve the equilibrium of physical parameters in a single-component fluid. These ingredients enable our scheme to achieve both high-order accuracy in the smooth region and the high resolution in the discontinuity region of the solution. Extensive numerical tests are performed to verify the high resolution and accuracy of the scheme.

关键词： finite volume high-order multicomponent flows

来源：评论

学校读者我要写书评

暂无评论

Correction: DC-PSENet: a novel scene text detection method integrating double ResNet-based and changed channels recursive feature pyramid

引用

The Visual Computer 2024年第2期41卷 1413-1414页

作者： Huang, Liwen Liao, Shujiao Yang, Wenyuan Fujian Key Laboratory of Granular Computing and Application Minnan Normal University Zhangzhou China School of Mathematics and Statistics Minnan Normal University Zhangzhou China

来源：评论

学校读者我要写书评

暂无评论

Secure Sharing Data Integrity Audit Scheme with Dynamic User Groups in Cloud Storage

Journal of Network Intelligence

引用

Journal of Network Intelligence 2024年第3期9卷 1673-1692页

作者： Liu, Jia-Xian Huang, Hui Chen, Qun-Shan Huang, Zhen-Jie School of Computer Science Minnan Normal University Zhangzhou363000 China Fujian Key Laboratory of Granular Computing and Application Minnan Normal University Zhangzhou363000 China

With the wide application of cloud storage, users with limited resources send outsourced data to the cloud service provider. Shared data storage has become a vital application form in some cloud storage scenarios. Many existing shared data integrity audit schemes don’t consider the user’s access rights. The revoked user can still access the data in the group. Moreover, they cannot support the dynamism of the data, nor can they support data privacy and user identity protection from infringement by the validator. We propose a safe and feasible shared data audit scheme to solve several of the problems above. In our scheme, once the group user is revoked, he has no permission to access and download data. We also design a dynamic multi-tree data storage structure to realize the increase, deletion, and search of dynamic user groups and data. Not only that, but our scheme also supports identity privacy protection for group users and data privacy protection. Experimental data show that our scheme is more efficient and safer than previous schemes. Especially in calculating tag generation, we reduces the computing overhead by 50 %. © 2024, Taiwan Ubiquitous Information CO LTD. All rights reserved.

关键词： Data integrity

来源：评论

学校读者我要写书评

暂无评论

Secure Location Data Sharing Scheme with KN-tree in Mobile Online Social Network(SMOSNs)

Journal of Network Intelligence

引用

Journal of Network Intelligence 2023年第2期8卷 609-626页

作者： Liu, Hong-Yang Huang, Hui Chen, Qun-Shan Huang, Zhen-Jie Wu, Chen-Huang School of Computer Science Minnan Normal University Zhangzhou363000 China School of Computer Science Fujian Key Laboratory of Granular Computing and Application Minnan Normal University Zhangzhou363000 China Fujian Key Laboratory of Granular Computing and Application Minnan Normal University Zhangzhou363000 China Key Laboratory of Applied Mathematics of Fujian Province University Putian University Putian351100 China

Location data is a critical foundation for online social networks, and sharing location data is essential to improve the efficiency of online social services. However, location data often contains users’ sensitive information. How to protect users’ privacy is a crucial problem in the process of location data sharing. Many solutions have been proposed to preserve user privacy, such as deploying trusted base stations or using en-cryption. But the problem of privacy disclosure still exists. In particular, the existing schemes do not consider that malicious third-party nodes can obtain the user’s location data through rainbow attacks. The location server can guess the social relationship of users through their frequent location requests. In this paper, we propose a brand-new scheme to ensure safety in the entire process. Salt encryption has been adopted to ensure data security in the location recording phase, and Bloom filter to ensure privacy security in the data sharing phase. Moreover, we proposed a KN-tree structure to simplify the computational process of location updates. Finally, the theoretical analysis shows that our scheme is superior to the existing methods in terms of security and practicality. We also realize and analyze the efficiency of our proposed scheme. And using the KN-tree structure to save users’ location data can reduce the computing cost of the location server by about 27%. © 2023, Taiwan Ubiquitous Information CO LTD. All rights reserved.

关键词： Social networking (online)

来源：评论

学校读者我要写书评

暂无评论

SOME GENERALIZED METRIC PROPERTIES OF n-SEMITOPOLOGICAL GROUPS

arXiv

引用

arXiv 2024年

作者： Lin, Fucai Qi, Xixi School of mathematics and statistics Minnan Normal University Zhangzhou363000 China Fujian Key Laboratory of Granular Computing and Application Minnan Normal University Zhangzhou363000 China

A semitopological group G is called an n-semitopological group, if for any g ∈ G with (Equation presented) there is a neighborhood W of e such that (Equation presented), where n ∈ N. The class of n-semitopological groups (n ≥ 2) contains the class of paratopological groups and Hausdorff quasi-topological groups. Fix any n ∈ N. Some properties of n-semitopological groups are studied, and some questions about n-semitopological groups are posed. Some generalized metric properties of n-semitopological groups are discussed, which contains mainly results are that (1) each Hausdorff first-countable 2-semitopological group admits a coarser Hausdorff semi-metrizable topology and has G∗δ-diagonal;(2) each locally compact, Baire and σ-compact 2-semitopological group is a topological group;(3) the condensation of some kind of 2-semitopological groups topologies are given. Finally, some cardinal invariants of n-semitopological groups are *** Codes Primary 54H11, 22A05, secondary 54A25, 54B15, 54E35 Copyright © 2024, The Authors. All rights reserved.

关键词： Topology

来源：评论

学校读者我要写书评

暂无评论

A decentralized data integrity auditing scheme for cloud storage

A decentralized data integrity auditing scheme for cloud sto...

引用

Data Security and Privacy Protection (DSPP), International Conference on

作者： Jiaxian Liu Hui Huang Qunshan Chen Zhenjie Huang School of Computer Science Minnan Normal University Zhangzhou Fujian China Fujian Key Laboratory of Granular Computing and Application Minnan Normal University Zhangzhou Fujian China

Cloud storage service providers offer remote data storage services, and users verify the integrity of the outsourced data through a fixed third-party auditor. However, a single third-party auditor may collaborate with other entities, which raises security problems. In this paper, we propose a decentralized integrity auditing scheme with multiple auditors participating, which solves the single point of failure problem caused by a single third-party auditor. However, using multiple auditors for auditing increases the communication overhead between cloud service providers and verifiers. To solve this problem, we introduce the concept of non-interactive public provable data ownership to reduce the interaction between them. We also design a new reward and punishment mechanism to motivate entities to participate honestly in auditing. The theoretical analysis and experimental evaluation show that our scheme is feasible and effective.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Consistency correction of echo intensity data for multiple radar systems and its application in quantitative estimation of typhoon precipitation

引用

Frontiers of Earth Science 2022年第1期16卷 99-108页

作者： Shuai ZHANG Jing HAN Bingke ZHAO Zhigang CHU Jie TANG Limin LIN Xiaoqin LU Jiaming YAN Shanghai Typhoon Institute China Meteorological AdministrationShanghai 200030Chin Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters Key Laboratory for Aerosol-Cloud-Precipitation(China Meteorological Administration)Nanjing University of Information Science&TechnologyNanjing 210044China Hainan Institute of Meteorological Sciences Haikou 570203China Key Laboratory of South China Sea Meteorological Disaster Prevention and Mitigation of Hainan Province Haikou 570203China Fujian Key Laboratory of Severe Weather Fuzhou 350001China Fujian Key Laboratory of Granular Computing and Application Minnan Normal UniversityZhangzhou 363000China

Calibration error is one of the primary sources of bias in echo intensity measurements by ground-based radar *** errors cause data discontinuity between adjacent radars and reduce the effectiveness of the radar *** Global Precipitation Measurement Kuband Precipitation Radar(GPM KuPR)has been shown to provide stable long-term *** this study,GPM KuPR observations were converted to S-band approximations,which were then matched spatially and temporally with ground-based radar *** measurements of stratiform precipitation below the melting layer collected by the KuPR during Typhoon Ampil were compared with those of multiple radar systems in the Yangtze River Delta to determine the deviations in the echo intensity between the KuPR and the ground-based radar *** echo intensity data collected by the ground-based radar systems was corrected using the KuPR observations as reference,and the correction results were verified by comparing them with rain gauge *** was found that after the correction,the consistency of the echo intensity measurements of the multiple radar systems improved significantly,and the precipitation estimates based on the revised ground-based radar observations were closer to the rain gauge measurements.

关键词： calibration error ground-based radar reflectivity correction precipitation estimates

来源：评论

学校读者我要写书评

暂无评论

Transversality on locally pseudocompact groups

引用

Frontiers of Mathematics in China 2021年第3期16卷 771-782页

作者： Fucai LIN Zhongbao TANG School of Mathematics and Statistics Minnan Normal UniversityZhangzhou363000China Fujian Key Laboratory of Granular Computing and Applications Minnan Normal UniversityZhangzhou363000China

Two non-discrete Hausdorff group topologiesτandδon a group G are called transversal if the least upper boundτ⋁δofτandδis the discrete *** this paper,we discuss the existence of transversal group topologies on locally pseudocompact,locally precompact,or locally compact *** prove that each locally pseudocompact,connected topological group satisfies central subgroup paradigm,which gives an affirmative answer to a problem posed by Dikranjan,Tkachenko,and Yaschenko[Topology Appl.,2006,153:3338–3354].For a compact normal subgroup K of a locally compact totally disconnected group G,if G admits a transversal group topology,then G/K admits a transversal group topology,which gives a partial answer again to a problem posed by Dikranjan,Tkachenko,and Yaschenko in ***,we characterize some classes of locally compact groups that admit transversal group topologies.

关键词： Transversal group topology locally pseudocompact group locally compact group locally precompact connected space central topological group

来源：评论

学校读者我要写书评

暂无评论

An anomaly detection approach based on hybrid differential evolution and K-means clustering in crowd intelligence

引用

International Journal of Crowd Science 2021年第2期5卷 129-142页

作者： Liu, Jianran Liang, Bing Ji, Wen Institute of Computing Technology Chinese Academy of Sciences Beijing China Institute of Computing Technology Chinese Academy of Sciences Beijing China Beijing Key Laboratory of Mobile Computing and Pervasive Device Institute of Computing Technology Chinese Academy of Sciences Beijing China Fujian Key Laboratory of Granular Computing and Application Minnan Normal University Zhangzhou China University of Chinese Academy of Sciences Beijing China

Purpose: Artificial intelligence is gradually penetrating into human society. In the network era, the interaction between human and artificial intelligence, even between artificial intelligence, becomes more and more complex. Therefore, it is necessary to describe and intervene the evolution of crowd intelligence network dynamically. This paper aims to detect the abnormal agents at the early stage of intelligent evolution. Design/methodology/approach: In this paper, differential evolution (DE) and K-means clustering are used to detect the crowd intelligence with abnormal evolutionary trend. Findings: This study abstracts the evolution process of crowd intelligence into the solution process of DE and use K-means clustering to identify individuals who are not conducive to evolution in the early stage of intelligent evolution. Practical implications: Experiments show that the method we proposed are able to find out individual intelligence without evolutionary trend as early as possible, even in the complex crowd intelligent interactive environment of practical application. As a result, it can avoid the waste of time and computing resources. Originality/value: In this paper, DE and K-means clustering are combined to analyze the evolution of crowd intelligent interaction. © 2020, Jianran Liu, Bing Liang and Wen Ji.

关键词： K-means clustering

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：