检索结果-内蒙古大学图书馆

您好，读者！请登录

内蒙古大学图书馆

首页
概况
党建
资源
服务
科研支持
- 论文收录引用证明
- 科技查新
知识产权
档案馆
帮助

咨询与建议

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

您的常用邮箱：*

您的手机号码：*

问题描述：

当前已输入0个字，您还可以输入200个字

全部搜索
期刊论文
图书
学位论文
标准
纸本馆藏
外文资源发现
数据库导航
超星发现

高级检索

时间限定

出版年份：

文献类型

图书期刊文献学位论文多媒体

馆藏选择

电子馆藏纸本馆藏

核心期刊

全部期刊 SCI 收录期刊 SSCI 收录期刊 EI 收录期刊 CSCD 收录期刊 CSSCI 收录期刊

语言

中文英文

文献类型

期刊文献图书学位论文标准纸本馆藏

帮助

文字说明：

T=题名（书名、题名），A=作者（责任者），K=主题词，P=出版物名称，PU=出版社名称，O=机构（作者单位、学位授予单位、专利申请人），L=中图分类号，C=学科分类号，U=全部字段，Y=年（出版发行年、学位年度、标准发布年）

检索规则说明：

AND代表“并且”；OR代表“或者”；NOT代表“不包含”；(注意必须大写,运算符两边需空一格)

检索范例：

范例一：(K=图书馆学 OR K=情报学) AND A=范并思 AND Y=1982-2016
范例二：P=计算机应用与软件 AND (U=C++ OR U=Basic) NOT K=Visual AND Y=2011-2016

分类表

所选分类

>> <<

限定检索结果

文献类型

3,157 篇 会议
72 篇 期刊文献
65 册 图书

馆藏范围

3,293 篇 电子文献
1 种 纸本馆藏

日期分布

学科分类号

2,338 篇 工学
- 2,058 篇 计算机科学与技术...
- 1,036 篇 软件工程
- 414 篇 电气工程
- 326 篇 信息与通信工程
- 310 篇 电子科学与技术（可...
- 112 篇 控制科学与工程
- 69 篇 机械工程
- 67 篇 光学工程
- 67 篇 生物工程
- 62 篇 生物医学工程（可授...
- 35 篇 动力工程及工程热...
- 33 篇 仪器科学与技术
- 32 篇 建筑学
- 30 篇 材料科学与工程（可...
- 29 篇 化学工程与技术
- 25 篇 土木工程
- 21 篇 力学（可授工学、理...
721 篇 理学
- 482 篇 数学
- 174 篇 物理学
- 79 篇 生物学
- 65 篇 系统科学
- 60 篇 统计学（可授理学、...
- 36 篇 化学
246 篇 管理学
- 158 篇 管理科学与工程(可...
- 102 篇 图书情报与档案管...
- 70 篇 工商管理
63 篇 医学
- 53 篇 临床医学
- 21 篇 基础医学(可授医学...
22 篇 农学
- 19 篇 作物学
21 篇 法学
- 19 篇 社会学
15 篇 经济学
12 篇 文学
11 篇 教育学
4 篇 军事学

主题

327 篇 parallel process...
204 篇 graphics process...
203 篇 computer archite...
157 篇 parallel archite...
136 篇 parallel process...
123 篇 parallel algorit...
121 篇 graphics process...
115 篇 hardware
113 篇 image processing
86 篇 concurrent compu...
86 篇 computational mo...
76 篇 signal processin...
72 篇 parallel program...
71 篇 field programmab...
68 篇 instruction sets
68 篇 multicore proces...
67 篇 parallel computi...
65 篇 algorithm design...
58 篇 throughput
57 篇 gpu

机构

9 篇 college of compu...
9 篇 natl univ def te...
8 篇 carleton univ sc...
8 篇 national laborat...
6 篇 hosei univ dept ...
6 篇 inria rennes
6 篇 st francis xavie...
5 篇 chinese acad sci...
5 篇 univ aizu dept c...
5 篇 polish japanese ...
5 篇 computer science...
5 篇 college of compu...
5 篇 city university ...
4 篇 shanghai jiao to...
4 篇 charles univ pra...
4 篇 rwth aachen univ...
4 篇 hainan internati...
4 篇 department of co...
4 篇 university of ch...
4 篇 universidad carl...

作者

11 篇 jack dongarra
10 篇 roman wyrzykowsk...
8 篇 dongarra jack
7 篇 liu jie
7 篇 konrad karczewsk...
7 篇 quintana-orti en...
6 篇 hannig frank
6 篇 li dongsheng
6 篇 teich juergen
6 篇 li chao
6 篇 nakano koji
6 篇 peng shietung
6 篇 li yamin
6 篇 chu wanming
6 篇 krulis martin
5 篇 zhang lei
5 篇 ito yasuaki
5 篇 li kenli
5 篇 wanlei zhou
5 篇 tudruj marek

语言

3,230 篇 英文
53 篇 其他
15 篇 中文

检索条件"任意字段=5th International Conference on Algorithms and Architectures for Parallel Processing"

共 3294 条记录，以下是591-600 订阅

全选清除本页清除全部题录导出标记到"检索档案"

详细简洁

排序：

Mapping YOLOv4-Tiny on FPGA-Based DNN Accelerator by Using Dynamic Fixed-Point Method

Mapping YOLOv4-Tiny on FPGA-Based DNN Accelerator by Using D...

引用

international Symposium on parallel architectures, algorithms and Programming (PAAP)

作者： Peng Li Cheng Che State Key Laboratory of Information Security Institute of Information Engineering Chinese Academy of Sciences Beijing China School of Cyber Security University of Chinese Academy of Sciences Beijing China

ISBN: (纸本)9781665496407

In the past few decades, with the large-scale application of deep learning technology, the neural network inference speed problem is becoming more and more severe, especially in some mobile and embedded real-time processing systems. To improve the inference speed, some software optimization techniques have been proposed by previous researchers. However, these methods only focused on the model self and still cannot meet the rapid development of large and complex deep learning models. Fortunately, deep neural network (DNN) accelerators based on FPGA SoC has opened a promising opportunity for the real-time inference. In this paper, we proposed a novel 16-bit dynamic fixed-point number quantization method to map the object detection network YOLOv4-tiny into FPGA-based heterogeneous deep learning accelerators. We evaluate this model in Xilinx Zynq-7020 SoC on the ZedBoard platform. Experiment on coco dataset shows that this approach can improve the model inference speed to 4× than pure CPU platform, and the accuracy loss is negligible with only 3-5%. this mapping method can be competitive with or even superior to other state-of-the-art accelerating method in deep learning fields.

关键词： Deep learning Quantization (signal) Neural networks Programming Real-time systems Software parallel architectures

来源：评论

学校读者我要写书评

暂无评论

Fast Radar-based Particle Filtering with Divergence Avoidance Strategies 5

Fast Radar-based Particle Filtering with Divergence Avoidanc...

引用

5th international conference on Control and Robotics Engineering (ICCRE)

作者： Pishehvari, Ahmad Lessmann, Stephanie Lategahn, Julian Tibken, Bernd Berg Univ Wuppertal Wuppertal Germany APTIV Serv Deutschland Gmbh Wuppertal Germany

ISBN: (纸本)9781728167916

In this paper the global self-localization problem of a vehicle without any prior knowledge of its initial pose in GNSS(1)-restricted outdoor areas is investigated. Particle filtering is applied using radars, mechanical motion sensors and a digital map. To prevent the filter divergence due to the noisy measurements, two novel clustering-based techniques are introduced. the first technique tracks the final generated clusters by their Gaussian parameters. the second approach processes all clusters as independent filters in parallel until the convergence conditions are fulfilled. To recognize the filter divergence and convergence for the first and second approach, respectively, the samples are continuously monitored by their effective sample size and entropy information. In order to filter recovery in case of divergence, a systematic particle injection is implemented for the clustering tracking technique which avoids a filter reset. To accelerate the filtering process, two novel additional map layers are generated by a pre-processing of the digital environment map, correlating with Voronoi diagram for line segments. the feasibility and validity of the proposed method is tested in a challenging static environment. the experimental results are verified qualitatively and quantitatively to evaluate the performance of the presented approach.

关键词： particle filtering global self-localization particle injection clustering parallel filtering cluster tracking

来源：评论

学校读者我要写书评

暂无评论

A Communication-Avoiding Algorithm for Molecular Dynamics Simulation 19th

A Communication-Avoiding Algorithm for Molecular Dynamics Si...

引用

19th international conference on algorithms and architectures for parallel processing (ICA3PP)

作者： Wang, Bei Chen, Yifeng Hou, Chaofeng Peking Univ EECS Beijing 100871 Peoples R China Chinese Acad Sci Inst Proc Engn Beijing Peoples R China

ISBN: (纸本)9783030389918;9783030389901

Molecular dynamics and many similar time-dependent computing tasks are defined as simple state updates over multiple time steps. In recent years, modern supercomputing clusters have enjoyed fast-growing compute capability and moderate-growing memory bandwidth, but their improvement of network bandwidth/latency is limited. In this paper, we propose a new communication-avoiding algorithmic model based on asynchronous communications which, unlike BSP, records and handles multiple iterative states together. the basic idea is to let computation run in small regular time steps while communications over longer dynamic time steps. Computation keeps checking inaccuracies so that the intervals between communications are small in volatile scenarios but longer when dynamics is smooth. this helps reduce the number of data exchanges via network communication and hence improve the overall performance when communication is the bottleneck. We test MD simulation of condensed covalent materials on the Sunway TaihuLight. For best time-to-solution, the general-purpose supercomputer Sunway TaihuLight performs 11.8K steps/s for a system with 2.1 million silicon atoms and 5.1 K steps/s for 50.4 million silicon atoms. this time-to-solution performance is close to those of state-of-art hardware solution. A software solution using general-purpose supercomputers makes the technology more accessible to the general scientific users.

关键词： BSP parallel-in-time Molecular dynamics

来源：评论

学校读者我要写书评

暂无评论

Shared Buffer Crossbar Architecture for GPU-CPU 5th

Shared Buffer Crossbar Architecture for GPU-CPU

引用

5th international conference on Microelectronics, Computing and Communication Systems, MCCS 2020

作者： Pitliya, Deepika Palecha, Namita Department of ECE R.V. College of Engineering Bangalore India

ISBN: (纸本)9789811602740

To achieve high-performance and energy optimized computing the GPU-CPU heterogeneous architectures are standout choice. Streaming multiprocessors (SMs) are increasing to boost throughput in GPUs. Design of on-chip interconnect for GPU-CPU diverse system is a challenge to make it scalable and efficient. Mesh network is being used in manycore CPUs but for GPU it consumes more area and power as well, due to traffic pattern of GPU. Crossbar is good fit, but it is not supporting communication among the SMs when numbers of SM are high. the motivation is to design the scalable crossbar which provide communication between SMs and SM to memory unit. the objective here to design two types of crossbar with shared buffer, crossbar local and global. Crossbar local provides communication among SMs and take all the input request which are going to the memory in coincide manner and pass it to crossbar global that divaricate these request to memory unit, Last-level cache as well as memory controllers. Sharing buffer give opportunity to all input for communication way efficient to achieve high throughput with reduce area and power. Compare to mesh network in Shared buffer crossbar network reduction in area 28% and power 32%. Scalability of design is verified by increasing the number of SMs. © 2021, the Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Graphics processing unit

来源：评论

学校读者我要写书评

暂无评论

Comparison of Image Encoder architectures for Image Captioning

Comparison of Image Encoder Architectures for Image Captioni...

引用

international conference on Computing Methodologies and Communication (ICCMC)

作者： Harsh Maru TSS Chandana Dinesh Naik NITK Karnataka India

ISBN: (纸本)9781665403603;9781665447751

Image captioning is a fascinating and challenging task combining two interesting fields in computer science - Natural Language processing and Computer Vision. Several approaches have been tried for this task. Many of these approaches are based on a combination of Convolutional Neural Networks(CNN's) as encoders and Recurrent Neural Networks(RNN's) as decoders. this paper compares two CNN-based encoders - (Vector Geometry Group) VGG16 and InceptionV3, which are used as encoders for encoding the input image features. the proposed research work uses the popular Flickr8k dataset and also compares among two different loss functions used with each of the above architectures- Categorical Cross Entropy loss function and Kullback-Leibler Divergence. Our main goal is to study the effect of the image encoder and the loss function on the image captioning task while keeping all other parameters the same.

关键词： Geometry Training Legged locomotion Recurrent neural networks Roads Computer architecture Entropy

来源：评论

学校读者我要写书评

暂无评论

Cross-shard Transaction processing in Sharding Blockchains 20th

Cross-shard Transaction Processing in Sharding Blockchains

引用

20th international conference on algorithms and architectures for parallel processing (ICA3PP)

作者： Liu, Yizhong Liu, Jianwei Yin, Jiayuan Li, Geng Yu, Hui Wu, Qianhong Beihang Univ Sch Elect & Informat Engn Beijing Peoples R China Beihang Univ Sch Cyber Sci & Technol Beijing Peoples R China Univ Copenhagen Dept Comp Sci Copenhagen Denmark

ISBN: (纸本)9783030602482;9783030602475

Sharding blockchains could improve the transaction throughput and achieve scalibility, making the application fields of the blockchain technology more extensive. Cross-shard transactions account for a large fraction of transactions in a sharding blockchain, so the processing method of cross-shard transactions is of vital importance to the system efficiency. In this paper, we focus on the study of cross-shard transaction processing methods. Firstly, a summary of cross-shard transaction processing methods for sharding blockchains is given. Secondly, we propose RSTBP, which is built on the basis of a two phase commit protocol. In RSTBP, an input shard runs an intra-shard consensus algorithm, i.e., a Byzantine fault tolerance (BFT) algorithm, to process multiple inputs of different transactions simultaneously. For each input, a corresponding proof of availability is generated and sent to the relevant shards. Compared with previous schemes, the number of BFT calls is reduced by hundreds of times when processing the same number of transactions. thirdly, RSTSBP is designed by making some modifications to RSTBP. the proofs of availability are constructed according to different shards. the Merkel tree structure is different from that of RSTBP to cut down message complexity of the proofs. Both of the two schemes are proved to satisfy the consistency, liveness and responsiveness properties, and improve the cross-shard transaction processing efficiency.

关键词： Sharding blockchains Byzantine fault tolerance Cross-shard transactions Scalability Responsiveness

来源：评论

学校读者我要写书评

暂无评论

Distributed and parallel Ensemble Classification for Big Data Based on Kullback-Leibler Random Sample Partition 20th

Distributed and Parallel Ensemble Classification for Big Dat...

引用

20th international conference on algorithms and architectures for parallel processing (ICA3PP)

作者： Wei, Chenghao Zhang, Jiyong Valiullin, Timur Cao, Weipeng Wang, Qiang Long, Hao Shenzhen Univ Big Data Inst Coll Comp Sci & Software Engn Shenzhen 518000 Peoples R China Hangzhou Dianzi Univ Sch Automat Hangzhou 311305 Peoples R China Southern Univ Sci & Technol SUSTech Acad Adv Interdisciplinary Studies Shenzhen 518055 Peoples R China

ISBN: (纸本)9783030602451;9783030602444

In this article, we use a Kullback-Leibler random sample partition data model to generate a set of disjoint data blocks, where each block is a good representation of the entire data set. Every random sample partition (RSP) block has a sample distribution function similar to the entire data set. To obtain the statistical measure between them, Kernel Density Estimation (KDE) with a dual-tree recursion data structure is firstly applied to fast estimate the probability density of each block. then, based on the Kullback-Leibler (KL) divergence measure, we can obtain the statistical similarity between a randomly selected RSP data block and other RSP data blocks. We rank the RSP data blocks according to their divergence values in descending order and choose the first ten for an ensemble classification learning. the classification models are established in parallel for the selected RSP data blocks and the final ensemble classification model is obtained by the weighted voting ensemble strategy. the experiments were conducted by building XGboost model based on those ten blocks in parallel, and we incrementally ensemble them according to their KL values. the testing classification results show that our method can increase the generalization capability of the ensemble classification model. It could reduce the model building time in parallel computation environment by using less than 15% of the entire data, which could also solve the memory constraints of big data analysis.

关键词： Big data analysis Approximate computing Random sample partition Ensemble classification parallel distributed computing

来源：评论

学校读者我要写书评

暂无评论

Classification Of Human Facial Expressions For Emotion Recognition Using A Distributed Computer System 5

Classification Of Human Facial Expressions For Emotion Recog...

引用

5th international conference on Computer Science and Engineering (UBMK)

作者： Korkmaz, Timucin Erol, Hamza Mersin Univ Grad Sch Nat & Appl Sci Dept Comp Engn Mersin Turkey Mersin Univ Engn Fac Dept Comp Engn Mersin Turkey

ISBN: (纸本)9781728175652

A computer system was developed to classify human facial expressions for emotion recognition using a distributed computer architecture within the scope of big data. Computers with normal standards and features were used in the distributed computer architecture. Necessary system software for the distributed computer architecture was established and mutual communication protocols were provided between the computers and databases in the computer network. Visual C# parallel programming was used as the software language on the distributed computer architecture. In the software prepared, face image files were processed and threads were created. the threads created later were processed in the processors of the computers. the threads were run on the processors in the distributed computer system, and facial expressions were classified for emotion recognition. In the distributed computer architecture: the number of image files, the volume of big data and the load of the threads to be processed arc taken into account;and databases and parallel programming were used and the classification of human face images was performed. Emotion analysis methods and facial expression recognition techniques were used in the classification process. the distributed computer system is in low cost and high processing speed. With the distributed computer system architecture, big data analytics has become convenient and feasible.

关键词： Big data Distributed System Emotion analysis Image classification parallel Programing Face expression recognition

来源：评论

学校读者我要写书评

暂无评论

Accelerating Nested Data parallelism: Preserving Regularity 26th

Accelerating Nested Data Parallelism: Preserving Regularity

引用

26th international conference on parallel and Distributed Computing (Euro-Par)

作者： van den Haak, Lars B. McDonell, Trevor L. Keller, Gabriele K. de Wolff, Ivo Gabe Eindhoven Univ Technol Eindhoven Netherlands Univ Utrecht Utrecht Netherlands

ISBN: (纸本)9783030576752;9783030576745

Irregular nested data-parallelism is a powerful programming model which enables the expression of a large class of parallel algorithms. However, it is notoriously difficult to compile such programs to efficient code for modern parallel architectures. Regular data-parallelism, on the other hand, is much easier to compile to efficient code, but too restricted to express some problems conveniently or in a manner to exploit the full parallelism. We extend the regular data-parallel programming model to allow for the parallel execution of array-level conditionals and iterations over irregular nested structures, and present two novel static analyses to optimise the code generated for these programs which reduces the costs of this more powerful irregular model. We present benchmarks to support our claim that these extensions are effective as well as feasible, as they enable to exploit the full parallelism of an important class of algorithms, and together with our optimisations lead to an improvement in absolute performance over an implementation limited to exploiting only regular parallelism.

关键词： Arrays GPGPU Nested data parallelism parallel functional programming

来源：评论

学校读者我要写书评

暂无评论

An Efficient High-Speed Lifting Based 1D/2D-DWT VLSI Architecture Using CDF-5/3 Wavelet Transform For Image processing Applications 5

An Efficient High-Speed Lifting Based 1D/2D-DWT VLSI Archite...

引用

5th IEEE international conference on Recent Trends in Electronics, Information and Communication Technology (RTEICT)

作者： Sushmitha, M. Chetan, S. Sarkar, Sayantam Dr Ambedkar Inst Technol Dept Elect & Commun Engn VLSI Design & Embedded Syst Bengaluru India Dr Ambedkar Inst Technol Dept Elect & Commun Engn Bengaluru India Vijaya Vittala Inst Technol Dept Elect & Commun Engn Bengaluru India

ISBN: (纸本)9781728197722

there are various Discrete Wavelet Transform architectures that are designed to fulfil certain requirements and criteria's. the convolution method which is an old traditional method which requires more multipliers, hardware resources and huge memory storage which is not apt to yield high speed and efficient image processing, signal processing application designs when compared to lifting method. In this paper, we have proposed an architecture for lifting scheme based CDF-5/3 2D-DWT, which includes less mathematical computations and high speed is achieved compared to existing works. the proposed architecture design is multiplier-less and utilizes lesser number of components hence it is area efficient and also achieves high throughput rate. the RTL of this design is described in VHDL language and is synthesized in Xilinx 14.5 and target device is Virtex-5 series XC5VLX110T for testing the resulting parameters like LUTs, flip- flops and slices. Higher PSNR values and lesser MSE values are achieved when tested for different kind of images. the maximum operating frequency achieved for the designed 2D-DWT architecture is 258.358MHz. Hence the proposed 2D-DWT architecture has increased speed, less computational complexity and good hardware utilization when compared to other existing works.

关键词： Lifting scheme CDF-5/3 Wavelet Filter DWT 1D-DWT 2D-DWT

来源：评论

学校读者我要写书评

暂无评论

没有更多数据了...

全选清除本页清除全部题录导出标记到“检索档案”

共330页 << < 56 57 58 59 60 61 62 63 64 65 > >>

检索报告对象比较合并检索0

隐藏清空

合并搜索

回到顶部

执行限定条件

内容：

评分：

请选择保存的检索档案：

请选择收藏分类：

订阅名称：

通借通还

温馨提示：

图书名称：

借书校区：

取书校区：

手机号码：

邮箱地址：

一卡通帐号：

电话和邮箱必须正确填写，我们会与您联系确认。

联系人：

所在院系：

联系邮箱：

联系电话：

内蒙古自治区呼和浩特市赛罕区大学西街235号邮编: 010021

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：