检索结果-内蒙古大学图书馆

International Joint Conference on Neural Networks (IJCNN)

作者： Xuesen Ma Jindian Dong Weixin Wei Biao Zheng Ji Ma Tianbao Zhou School of Computer Science and Information Engineering Hefei University of Technology Hefei China Intelligent Interconnected Systems Laboratory of Anhui Province Hefei University of Technology Hefei China

Remote sensing technology is becoming more sophisticated and is extensively used for object tracking, urban planning, military reconnaissance and other fields. Complex backgrounds and diverse object scales are two important factors that affect the object detection effect of remote sensing images. To address this problem, this paper proposes a remote sensing object detection model that incorporates channel enhancement and multi-scale contextual features. Firstly, the multi-scale contextual feature enhancement module is constructed, which performs multi-order spatial interaction by cascading recursive convolution to obtain contextual information of feature maps at different scales, and introduces attention to reinforce unique features of objects and suppress background interference. Then, the spatial pyramid channel enhancement module combining sub-pixel convolution and adaptive sampling factor is designed to mitigate the semantic weakening of the depth feature maps caused by channel downscaling, thus enhancing the sampling effect between feature maps of different scales and reducing information loss. Finally, the effectiveness of the model is verified on the large-scale remote sensing image object detection dataset DIOR.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Visualizing Student Attention in Smart Classrooms: A Preliminary Study

Visualizing Student Attention in Smart Classrooms: A Prelimi...

引用

International Conference on Innovations in information Technology

作者： Medha Mohan Ambali Parambil Amna Mohammed Abdulla Alhammadi Fady Alnajjar Zouheir Trabelsi Luqman Ali Muhammed Swavaf Department of Information Systems and Security College of IT United Arab Emirate University Al Ain UAE Department of Computer Science and Software Engineering College of IT United Arab Emirate University Al Ain UAE

In recent years, there has been a swift progression in employing novel methods in classrooms to enhance students’ academic achievements, especially in line with the growing digitization of education. Such methods often encompass systems like facial recognition to monitor various aspects of students, including attendance, emotional states, and attention. These tools are capable of evaluating students’ presence and engagement in class, offering quantifiable metrics regarding their concentration and emotions. However, a prominent challenge has been the translation of this data into an accessible form that enables educators to assess and enhance their teaching techniques swiftly. Our suggested solution tackles this issue by offering a real-time visual depiction of students’ classroom status through different visualization techniques. These visual aids allow teachers to promptly recognize trends in student focus, thus aiding in the strategic alteration of teaching styles. Furthermore, these visual representations can be tailored to display various metrics and applied to tasks beyond monitoring attention, like overseeing attendance or assessing student progress. By integrating these advanced visualizations into the educational process, both teaching efficacy and the learning experience for students and teachers alike can be substantially elevated.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Scaling Law Based D2D Wireless Ad-Hoc Networks in the Finite Blocklength Regime

Scaling Law Based D2D Wireless Ad-Hoc Networks in the Finite...

引用

IEEE Global Communications Conference

作者： Xi Zhang Qixuan Zhu H. Vincent Poor Department of Electrical and Computer Engineering Networking and Information Systems Laboratory Texas A&M University College Station TX USA Department of Electrical Engineering Princeton University Princeton NJ USA

ISBN: (数字)9781728109626

ISBN: (纸本)9781728109633

A device-to-device (D2D) wireless ad hoc network architecture enables dynamic self-organizing communications among mobile users who can directly exchange information with their peers without a pre-determined network infrastructure. Moreover, finite blocklength coding (FBC) is the promising candidate technique to support time sensitive multimedia wireless networks services, where mobile users transmit short packets to upper-bound the transmission delay of video/audio traffic. The scaling law technique models the maximum D2D channel capacity as a function of the density of mobile users. Recent studies have integrated D2D wireless ad hoc networks with FBC theory to further improve the performance of 5G wireless ad hoc networks. However, how to model and analyze the capacity of D2D wireless ad hoc networks under the finite blocklength regime is not well understood and has not been thoroughly studied. To overcome these challenges, applying the scaling law technique, we derive upper-bounds on the coding rate of each D2D channel and the number of time slots needed to complete all D2D transmissions. Combining the D2D channel's coding rate with the number of time slots needed for all D2D transmissions, we derive the maximum aggregate throughput for wireless ad hoc networks with all mobile users using D2D communications while mitigating interference. We also develop a model where each D2D channel follows the Nakagami-m distribution, under which we derive the average aggregate throughput and its upper-bound. Finally, we evaluate our derived results in the D2D wireless ad hoc networks over finite blocklength regime through numerical analyses.

关键词： Device-to-device communication Mobile ad hoc networks Encoding Throughput Wireless communication Aggregates

来源：评论

学校读者我要写书评

暂无评论

Scaling-Up Item-Based Collaborative Filtering Recommendation Algorithm Based on Hadoop

Scaling-Up Item-Based Collaborative Filtering Recommendation...

引用

IEEE World Congress on Services (SERVICES)

作者： Jing Jiang Jie Lu Guangquan Zhang Guodong Long Decision Systems & E-Service Intelligence Research Laboratory Center of Quantum Computing and Intelligent System School of Software Faculty of Engineering and Information Technology Sydney University of Technology Sydney NSW Australia

Collaborative filtering (CF) techniques have achieved widespread success in E-commerce nowadays. The tremendous growth of the number of customers and products in recent years poses some key challenges for recommender systems in which high quality recommendations are required and more recommendations per second for millions of customers and products need to be performed. Thus, the improvement of scalability and efficiency of collaborative filtering (CF) algorithms become increasingly important and difficult. In this paper, we developed and implemented a scaling-up item-based collaborative filtering algorithm on MapReduce, by splitting the three most costly computations in the proposed algorithm into four Map-Reduce phases, each of which can be independently executed on different nodes in parallel. We also proposed efficient partition strategies not only to enable the parallel computation in each Map-Reduce phase but also to maximize data locality to minimize the communication cost. Experimental results effectively showed the good performance in scalability and efficiency of the item-based CF algorithm on a Hadoop cluster.

关键词： Scalability Prediction algorithms Algorithm design and analysis Collaboration Filtering Partitioning algorithms Clustering algorithms

来源：评论

学校读者我要写书评

暂无评论

Data-driven protocol off-loading for Ad hoc networking environment

Data-driven protocol off-loading for Ad hoc networking envir...

引用

2008 International Conference on Parallel and Distributed Processing Techniques and Applications, PDPTA 2008

作者： Aoki, Kazuhiro Ishii, Hiroshi Mizuno, Osamu Iwata, Makoto Nishikawa, Hiroaki Information Infrastructure Laboratory Inc. Saiki 210-4 Tsukuba Ibaraki 300-0028 Japan School of Information Science and Technology Tokai University Kitakaname 1117 Hiratsuka Kanagawa 259-1292 Japan NTT Information Sharing Platform Laboratories Midori-cho 3-9-11 Musashino Tokyo 180-8585 Japan Department of Information System Engineering Kochi University of Technology Tosayamada-cho Miyanokuchi 185 Kami Kochi 782-8502 Japan Department of Computer Science Graduate School of Systems and Information Engineering University of Tsukuba Ten'nodai 1-1-1 Tsukuba Ibaraki 305-8573 Japan

ISBN: (纸本)1601320841

Pervasive networking environment as social infrastructure has been going to become a reality, especially through the recent considerable progress of semiconductor technology. Considering the case where next generation of pervasive networking is realized over ad hoc network suitable for emergency and some tentative event, some ofauthors study about an data- driven imprementation of Ad hoc and ubiquitous communication environment. It is very important to implement effective protocol handling and routing process in Ad hoc networking environment especially. This paper describes the effectiveness ofan implementation ofprotocol off-loader using networking-oriented data-driven processors CUE (Coordinating User's requirements and engineering constraints). In Ad hoc network, it is necessary to realize connection-less protocol such as UDP/IPfor flexible routing and realtime communication. Firstly, this paper discusses data- driven implementation ofprotocol off-loader on CUE networking board to minimize overheads in protocol handling. Then this paper describes architecture of CUE-v2/CUE-v3 which is available version in CUE processors. This paper finally discusses effectiveness ofprotocol handling off-loader using CUE processor system. It shows that data-driven protocol off-loader can keep minimum turn-around time without runtime overheads.

关键词： Ad hoc networks

来源：评论

学校读者我要写书评

暂无评论

4.4.1 Fine-grained Method and Tool Integration for better Automotive software

引用

INCOSE International Symposium 2006年第1期16卷 606-620页

作者： Altheide, Frank Buhr, Kerstin Doerr, Heiko University of Paderborn Workgroup Database and Information Systems Germany Technical University of Berlin Institute for Software Engineering and Theoretical Computer Science Germany DaimlerChrysler AG Research & Technology Department Methods & Tools Germany

In the automotive field, software development methods and tools are used to cope with the high complexity of automotive software development. However, problems occur with the tracing of information, the assessment and monitoring of the development status, late and thus expensive changes, etc, which lead to increasing development costs and time as well as quality deficits. Our experience with recent projects have shown that the key to solving these problems lies in providing proper answers to the following questions: (1) How can development methods be assembled? (2) How can the use of a variety of development methods be supported by an appropriate tool?. To answer these questions, we introduce a tool integration approach that is strongly based on the integration of development methods. The contribution of this paper is (a) a method integration approach that considers finegrained method objects and their specific structure in addition to the usually considered artifacts, (b) an approach to tool integration that overcomes the deficits of heterogeneous tool environments and utilizes the results of the methodical understanding. © 2006 The Authors.

关键词： Integration

来源：评论

学校读者我要写书评

暂无评论

Cyclegan-Based SAR Image Expansion to Enhance Target Detection Performance

Cyclegan-Based SAR Image Expansion to Enhance Target Detecti...

引用

IEEE International Symposium on Geoscience and Remote Sensing (IGARSS)

作者： Fang Zhou Rong Liu Tingting Yang Xiyu Yang School of Computer Science and Information Engineering Hefei University of Technology Hefei China The Intelligent Interconnected Systems Laboratory of Anhui Province Hefei University of Technology Hefei China

ISBN: (数字)9798350360325

ISBN: (纸本)9798350360332

In response to the challenge of limited training samples for synthetic aperture radar (SAR) target detection, a SAR image expansion method based on improved cycle generative adversarial network (CycleGAN) is proposed. First, the structure of the CycleGAN is refined to enhance training efficiency. Then, instance normalization (IN) is incorporated into the residual unit of the CycleGAN to improve the realism of generated SAR images. Finally, SAR target recognition experiments were conducted using the single shot multibox detection (SSD) detection network with VoVNet structure. The experimental results prove that the data expansion method based on CycleGAN can improve the recognition accuracy of SAR targets.

关键词： Training Accuracy Target recognition Geoscience and remote sensing Object detection Generative adversarial networks Radar polarimetry

来源：评论

学校读者我要写书评

暂无评论

Variable precision multigranulation rough set and attributes reduction

Variable precision multigranulation rough set and attributes...

引用

2012 Joint Rough Set Symposium, JRS 2012

作者： Ju, Hengrong Yang, Xibei Dou, Huili Song, Jingjing School of Computer Science and Engineering Jiangsu University of Science and Technology Zhenjiang 212003 Jiangsu China Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information Nanjing University of Science and Technology Ministry of Education Nanjing 210094 China Artificial Intelligence Key Laboratory of Sichuan Province Zigong 643000 China School of Economics and Management Nanjing University of Science and Technology Nanjing 210094 China

ISBN: (纸本)9783662446799

Multigranulation rough set is a new expansion of the classical rough set since the former uses a family of the binary relations instead of single one for the constructing of approximations. In this paper, the model of the variable precision rough set is introduced into the multigranulation environment and then the concept of the variable precision multigranulation rough set is proposed, which include optimistic and pessimistic cases. Not only basic properties of variable precision multigranulation rough set are investigated, but also the relationships among variable precision rough set, multigranulation rough set and variable precision multigranulation rough set are examined. Finally, a heuristic algorithm is presented for computing reducts of variable precision multigranulation rough set, it is also tested on five UCI data sets. © 2014 Springer-Verlag.

关键词： Heuristic algorithms

来源：评论

学校读者我要写书评

暂无评论

mmWave-MIMO Based 5G Wireless Ad-Hoc Networks in the Finite Blocklength Regime

mmWave-MIMO Based 5G Wireless Ad-Hoc Networks in the Finite ...

引用

2019 IEEE Global Communications Conference (GLOBECOM)

作者： Xi Zhang Jingqing Wang H. Vincent Poor Department of Electrical and Computer Engineering Networking and Information Systems Laboratory Texas A&M University College Station TX USA Department of Electrical Engineering Princeton University Princeton NJ USA

ISBN: (数字)9781728109626

ISBN: (纸本)9781728109633

The integration of millimeter wave (mmWave) and multiple-input and multiple-output (MIMO) techniques has been designed to provide reliable communications with large degrees of freedom while supporting the explosively growing number of mobile users. Under stringent requirements in terms of latency and reliability, due to the infinite blocklength assumption of the Shannon's capacity result, researchers have investigated new methods to characterize wireless data transmissions considering the block error probability. The finite blocklength coding (FBC) technique has been developed to model the finite blocklength coding rate in the non-asymptotic regime while supporting short-packet communications over 5G wireless ad-hoc networks. However, because of the design complexity when characterizing the second-order coding rate over mmWave MIMO based wireless channels while being integrated with FBC, how to accurately derive the finite blocklength coding rate over mmWave MIMO wireless fading channels is still an open problem over 5G wireless ad-hoc networks. To tackle the above-mentioned challenges, we propose and develop a system model that can efficiently integrate mmWave-MIMO techniques with finite blocklength coding over 5G wireless ad-hoc networks. In particular, we derive system equations that characterize the foundational informationtheoretic relationship between the finite blocklength channel capacity and the coding rate over our proposed mmWave MIMO based 5G wireless ad-hoc networks in the finite blocklength regime. Also conducted is a MATLAB-based performance evaluation, which validates and analyzes our proposed schemes over mmWave MIMO based 5G wireless ad-hoc networks in the finite blocklength regime.

关键词： Ad hoc networks Wireless communication MIMO communication 5G mobile communication Channel coding Mathematical model

来源：评论

学校读者我要写书评

暂无评论

Feasibility conditions of interference alignment with general alignment set

Feasibility conditions of interference alignment with genera...

引用

Asilomar Conference on Signals, systems & computers

作者： Liangzhong Ruan Vincent K.N. Lau Moe Z. Win Department of Electronic and Computer Engineering Hong Kong University of Science and Technology Hong Kong China Laboratory for Information and Decision Systems (LIDS) Massachusetts Institute of Technology USA

Interference alignment (IA) has received great recent attention for its breakthrough performance. Classical IA algorithms require infinite dimensional symbol extension. On the other hand, IA algorithms which draw on the finite signal dimension provided by multiple antennas, become infeasible when the network scales. To construct practical IA algorithms that are applicable to large-scale MIMO interference networks, it is essential to introduce flexible IA algorithms which selectively cancels the strongest interference perceived by the receivers (Rxs). In this work, we aim at understanding the feasibility conditions of such IA algorithms, which cancel interference selectively according to the alignment set. By exploiting methodologies from algebraic geometry and graph theory, we characterize the feasibility regions of IA schemes with general alignment set and obtain insights into how network configurations and alignment set affects the feasibility of IA schemes.

关键词：

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：