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

18,700 篇 会议
1,406 篇 期刊文献
713 册 图书
59 篇 学位论文
1 篇 科技报告

馆藏范围

20,880 篇 电子文献
4 种 纸本馆藏

日期分布

学科分类号

12,292 篇 工学
- 9,562 篇 计算机科学与技术...
- 4,627 篇 软件工程
- 3,225 篇 电气工程
- 2,846 篇 信息与通信工程
- 1,588 篇 生物工程
- 1,409 篇 控制科学与工程
- 1,206 篇 光学工程
- 1,141 篇 生物医学工程（可授...
- 710 篇 电子科学与技术（可...
- 548 篇 机械工程
- 431 篇 仪器科学与技术
- 298 篇 化学工程与技术
- 296 篇 材料科学与工程（可...
- 240 篇 网络空间安全
- 237 篇 安全科学与工程
- 220 篇 动力工程及工程热...
4,759 篇 理学
- 2,312 篇 物理学
- 1,759 篇 生物学
- 1,435 篇 数学
- 446 篇 统计学（可授理学、...
- 372 篇 化学
- 245 篇 系统科学
1,815 篇 医学
- 1,551 篇 临床医学
- 598 篇 基础医学(可授医学...
- 363 篇 药学(可授医学、理...
1,573 篇 管理学
- 947 篇 管理科学与工程(可...
- 706 篇 图书情报与档案管...
- 240 篇 工商管理
216 篇 农学
- 193 篇 作物学
201 篇 教育学
- 195 篇 教育学
140 篇 法学
61 篇 经济学
42 篇 文学
25 篇 军事学
14 篇 艺术学
3 篇 哲学
2 篇 历史学

主题

3,530 篇 neural networks
2,894 篇 artificial neura...
2,442 篇 convolutional ne...
1,841 篇 deep learning
1,570 篇 feature extracti...
1,485 篇 image processing
1,428 篇 training
1,332 篇 artificial intel...
914 篇 image segmentati...
911 篇 accuracy
696 篇 machine learning
619 篇 deep neural netw...
604 篇 neurons
574 篇 computational mo...
461 篇 image recognitio...
451 篇 image classifica...
415 篇 computer archite...
413 篇 computer vision
402 篇 signal processin...
384 篇 object detection

机构

48 篇 chitkara univers...
27 篇 univ chinese aca...
23 篇 tsinghua univers...
20 篇 school of comput...
19 篇 university of el...
19 篇 university of ch...
18 篇 university of sc...
18 篇 poznan univ life...
18 篇 nanyang technolo...
17 篇 shanghai jiao to...
16 篇 xidian univ sch ...
15 篇 school of comput...
15 篇 school of electr...
14 篇 hendisli&#x011f
14 篇 tsinghua univ pe...
13 篇 tongji universit...
13 篇 institute of aut...
13 篇 school of comput...
13 篇 beijing institut...
12 篇 shanghai jiao to...

作者

19 篇 de-shuang huang
16 篇 ryszard tadeusie...
16 篇 marcin korytkows...
15 篇 jacek m. zurada
11 篇 madani kurosh
11 篇 li yang
11 篇 liu yang
11 篇 liu ying
11 篇 ngoc thanh nguye...
10 篇 wang lei
10 篇 wang jing
10 篇 jun wang
9 篇 li xin
9 篇 haijun zhang
9 篇 gill kanwarparta...
9 篇 tianyong hao
9 篇 jun zhang
9 篇 li li
9 篇 shuzhi sam ge
9 篇 kok wai wong

语言

19,066 篇 英文
1,472 篇 其他
242 篇 中文
174 篇 土耳其文
7 篇 德文
5 篇 西班牙文
4 篇 法文
3 篇 俄文
2 篇 葡萄牙文
1 篇 日文

检索条件"任意字段=Conference on Applications of Artificial Neural Networks in Image Processing"

共 20880 条记录，以下是4961-4970 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：

SroX: Accelerating Hidden Weight Creation in neural networks via Spectral Relaxation and X-Iteration

SroX: Accelerating Hidden Weight Creation in Neural Networks...

引用

Advanced Computing & Communication Technologies (ICACCTech), International conference on

作者： Harshit Agarwal Parth Middha Ayush Thakur Sofia Singh Shipra Saraswat Neha Bhatia Dept. of Computer Science & Engineering Amity University Noida India Amity Institute of Information Technology Amity university Noida India Dept. of Information Technology Panipat Institute of Engineering & Technology Panipat India

ISBN: (数字)9798331519056

ISBN: (纸本)9798331519063

This paper proposes SroX, a novel approach for accelerating Hidden Weight creation in neural networks, which is a critical component of neural network training that can significantly impact the overall efficiency of the training process. SroX leverages spectral relaxation and X-Iteration to reduce the computational time required for Hidden Weight creation, enabling faster training times without sacrificing accuracy. Our approach addresses the long-standing challenge of efficient Hidden Weight creation, which has been a major bottleneck in neural network training. We demonstrate the efficacy of SroX through extensive experiments on benchmark datasets, including MNIST, CIFAR-10, and imageNet, where SroX outperforms existing methods, including Random Projection and Gradient-Based Optimization, in terms of training time and accuracy. The proposed approach provides a promising solution for large-scale neural network applications, where efficient training is crucial, and has the potential to revolutionize the field of neural network research and applications, driving innovation and breakthroughs in areas such as computer vision, natural language processing, and beyond. By providing a faster and more efficient way to create Hidden Weights, SroX can enable researchers and practitioners to tackle complex problems with unprecedented speed and efficiency, leading to significant advancements in the field of artificial intelligence.

关键词： Training Technological innovation Computer vision Accuracy Scalability neural networks Benchmark testing Natural language processing Communications technology Optimization

来源：评论

学校读者我要写书评

暂无评论

Scalable CNN Accelarator with Error-Correcting Accumulator for High-Performance Deep Learning applications

Scalable CNN Accelarator with Error-Correcting Accumulator f...

引用

Distributed Systems, Computer networks and Cybersecurity (ICDSCNC), International conference on

作者： Usthulamuri Penchalaiah G. Suresh K. Sravan Kumar G. Tejaswi A. Hari Krishna Dept. of ECE Geethanjali Institute of Science and Technology S.P.S.R. Nellore India

ISBN: (数字)9798350375442

ISBN: (纸本)9798350375459

The Multiply and Accumulate (MAC) unit is a fundamental component in Convolutional neural Network (CNN) accelerators, designed to enhance the efficiency and speed of deep learning applications such as image recognition, natural language processing, and autonomous driving. These applications demand high computational power and low latency, which are critical for real-time processing and decision-making. Existing solutions often struggle with power consumption, area efficiency, and processing speed, hindering their performance in resource-constrained environments. The proposed CNN-MAC accelerator addresses these challenges by integrating a Modified Booth Multiplier (MBM) and an Error Correctable Carry Look Ahead Adder (EC-CLA). The MBM enhances multiplication efficiency by reducing the number of partial products, while the EC-CLA minimizes propagation delay and corrects errors during addition, ensuring high-speed and reliable computation. This combination significantly improves the overall performance and energy efficiency of CNN accelerators, making them suitable for deployment in various high-demand scenarios.

关键词： Deep learning Power demand Resistive RAM Memory management Real-time systems Energy efficiency Computational efficiency Convolutional neural networks Low latency communication Pipeline processing

来源：评论

学校读者我要写书评

暂无评论

Medical image Segmentation Using Deep neural networks 29

Medical Image Segmentation Using Deep Neural Networks

引用

29th IEEE conference on Signal processing and Communications applications (SIU)

作者： Aydin, Musa Fatih Sultan Mehmet Vakif Univ Bilgisayar Muihendisligi Istanbul Turkey

ISBN: (纸本)9781665436496

Automatic segmentation of medical images plays an important role in diagnosing various diseases using these images. image segmentation can be applied separately through many medical samples to make different analyses and diagnoses. For example, by auto-segmentation cells in a cell culture, the amount, vitality, diameter or shape of the cells in the cell culture can be analysed separately. In addition, by automatic segmentation of blood vessels in a tissue, analyses such as the length and density of the vessels, and the radius of each vessel can be made separately. In this way, inferences such as early diagnosis of a disease and the type of the disease can be made from the examined image. In this study, using two different data sets consisting of various cell images and retinal images, the images in this data set were segmented separately. For automatic segmentation, a convolutional neural network model U-Net is used. The existing ground truth images and the images segmented using the U-Net network were compared. Percentage accuracy values obtained for both data sets, respectively;95.2 for retina data set and 97.25 for cell data set.

关键词： image Segmentation Deep neural networks Blood Vessel Segmentation Cell Nuclei Segmentation

来源：评论

学校读者我要写书评

暂无评论

Unsupervised Hyperspectral Anomaly Detection with Convolutional neural networks 29

Unsupervised Hyperspectral Anomaly Detection with Convolutio...

引用

29th IEEE conference on Signal processing and Communications applications (SIU)

作者： Yilmaz, Fatma Nur Arisoy, Sertac Kayabol, Koray Gebze Tekn Univ Elekt Muhendislgi Bolumu Kocaeli Turkey

ISBN: (纸本)9781665436496

In this study, we propose a convolutional neural network-based method for unsupervised hyperspectral anomaly detection. We obtain within- and between-cluster pixel pairs from hyperspectral images using the cluster maps obtained by an automatic clustering algorithm and constitute training data set by taking the differences of the pixel pairs. We design a multilayer convolutional neural network and train it using difference vectors. In prediction step, we apply a dual local window to hyperspectral image. For each pixel, we calculate the difference vectors between the center pixel and the surrounding pixels. By feeding the trained network with the difference vectors, we obtain prediction scores. In the last step, a weighted RX detector was obtained using prediction scores and used for anomaly detection. It has been observed that the experimental results conducted on four different real hyperspectral data show better results than the current CNN-based anomaly detector.

关键词： Hyperspectral image anomaly detection deep learning convolutional neural networks weighted RX

来源：评论

学校读者我要写书评

暂无评论

Temporal-Coded Deep Spiking neural Network with Easy Training and Robust Performance 35

Temporal-Coded Deep Spiking Neural Network with Easy Trainin...

引用

35th AAAI conference on artificial Intelligence / 33rd conference on Innovative applications of artificial Intelligence / 11th Symposium on Educational Advances in artificial Intelligence

作者： Zhou, Shibo Li, Xiaohua Chen, Ying Chandrasekaran, Sanjeev T. Sanyal, Arindam SUNY Binghamton Dept ECE Binghamton NY USA Harbin Inst Technol Harbin Peoples R China Univ Buffalo Dept EE Buffalo NY USA

ISBN: (纸本)9781577358664

Spiking neural network (SNN) is promising but the development has fallen far behind conventional deep neural networks (DNNs) because of difficult training. To resolve the training problem, we analyze the closed-form input-output response of spiking neurons and use the response expression to build abstract SNN models for training. This avoids calculating membrane potential during training and makes the direct training of SNN as efficient as DNN. We show that the non-leaky integrate-and-fire neuron with single-spike temporalcoding is the best choice for direct-train deep SNNs. We develop an energy-efficient phase-domain signal processing circuit for the neuron and propose a direct-train deep SNN framework. Thanks to easy training, we train deep SNNs under weight quantizations to study their robustness over low-cost neuromorphic hardware. Experiments show that our direct-train deep SNNs have the highest CIFAR-10 classification accuracy among SNNs, achieve imageNet classification accuracy within 1% of the DNN of equivalent architecture, and are robust to weight quantization and noise perturbation.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Smart Face Recognition Application using Deep Learning neural networks with Compressed Model Deployment

Smart Face Recognition Application using Deep Learning Neura...

引用

Applied artificial Intelligence and Computing (ICAAIC), International conference on

作者： Raj Kumar Bathina Raghu Ram Edara Bommalavarapu Gopi Jagadeesh Perugu Vegesna Venkatasiva Naga Yaswanth Lakshmi Narayana Thalluri Department of ECE Andhra Loyola Institute of Engineering and Technology Vijayawada A.P. India

This research work presents a novel face recognition system with customized face dataset. After completion of the dataset, this dataset will be fed to neural network (or) model and find how much efficient it is in detecting the student. There are different types of neural networks like ANN and its variants like CNN and RNN. For example, a person is affected by fever or cold etc. In that case, through biometric application disease the can be spread from one person to the other persons. To combat this situation, image processing authentication technique is proposed. This research study develops a new html script for accessing the camera to capture images. The input image is then fed to neural network. Here, a customized CNN model is developed. After recognition, the embedding or features of the image will be compared with the dataset. The proposed model will act like a black box, where input must be provided and it will process the image data and provide the recognized output.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Deep-BMNN: Implementing Sparse Binary neural networks in Memory-Based Reconfigurable Processor (MRP)

Deep-BMNN: Implementing Sparse Binary Neural Networks in Mem...

引用

International Technical conference o Circuits/Systems, Computers and Communications (ITC-CSCC)

作者： Kenta Sasagawa Senling Wang Tetsuya Nishikawa Hiroshi Kai Yoshinobu Higami Hiroshi Takahashi Hiroyuki Yotsuyanagi Tianming Ni Xiaoqing Wen Ehime University Tokushima University Anhui Polytechnic University Kyushu Institute of Technology

ISBN: (数字)9798350379051

ISBN: (纸本)9798350379068

The Memory-based Reconfigurable Processor (MRP) is a memory-centric reconfigurable device that offers advantages in processing speed and cost compared to other reconfigurable devices like FPGAs, making it a promising alternative device for artificial Intelligence applications such as neural networks (NNs). However, due to the unique structure of MRP, the method for implementing NNs on it has not yet been established. In our previous work [8], we have proposed a sparse neural network model named MNN (MRP neural networks), specifically designed to align with the unique connection structure of MRP for neural network integration. This paper extends our research by proposing a binarization method for MNN referred to as the BMNN using XNOR-Net to facilitate practical NNs application (handwriting digits recognition) on MRP. To enhance the performance of BMNN, a novel implementation strategy Deep-BMNN is also proposed aimed at increasing both the layer depth and the neuron count of BMNN by folding the logic routing on the MRP device. Experimental results using MNIST dataset for handwritten digits recognition demonstrate that our proposed binary NNs for MRP can achieve comparable accuracy to those of fully connected NNs employing real numbers.

关键词： Performance evaluation Handwriting recognition Accuracy Reconfigurable devices Neurons artificial neural networks Routing

来源：评论

学校读者我要写书评

暂无评论

A Superscalar Six-Stage Pipeline neural Network Accelerator: Design and Implementation Based on FPGA

A Superscalar Six-Stage Pipeline Neural Network Accelerator:...

引用

IEEE International conference on artificial Intelligence and Computer applications (ICAICA)

作者： Shaoshan Guo Yunqian Huang Kailan Mai Guopeng Mo Guangdong University of Technology Guangdong China

ISBN: (数字)9798350377255

ISBN: (纸本)9798350377262

The growing need for efficient neural network inference in embedded systems has spurred the development of specialized hardware accelerators. This paper introduces the design and implementation of an FPGA-based neural processing Unit (NPU), optimized for high-performance neural network tasks. The NPU architecture integrates advanced features, including SIMD (Single Instruction, Multiple Data) support, an out-of-order execution system, and a six-stage pipeline, ensuring efficient processing. The NPU's RAM system manages both instruction and data storage, while 16-bit floating-point (FP16) computations balance precision with resource efficiency. On the software side, the NPU utilizes a custom instruction set, supported by a C language framework and a Python-based compiler. Detailed analyses of key modules, such as scalar and matrix operations, neural network inference, and RAM storage, are provided. Simulation tests, including handwritten digit recognition using the MNIST dataset, validate the NPU's functionality and performance, demonstrating its robust computational capabilities and efficiency. This study offers a novel and practical approach to achieving efficient neural network inference on FPGA, underscoring the NPU's superior performance in handling complex tasks.

关键词： Out of order Software design Single instruction multiple data neural networks Pipelines Random access memory C languages Software Computational efficiency Field programmable gate arrays

来源：评论

学校读者我要写书评

暂无评论

Lightweight Text Matching Method with Rich Features 22

Lightweight Text Matching Method with Rich Features

引用

8th International conference on Computing and artificial Intelligence, ICCAI 2022

作者： Ji, Changhua Zhang, Tao Mao, Jiayi Zhang, Li School of Artificial Intelligence and Computer Science Jiangnan University Wuxi214122 China

ISBN: (纸本)9781450396110

Text matching is one of the research hotspots in Natural Language processing (NLP). The study of text matching is of great practical importance for applications such as text de-duplication, web retrieval, and question answering systems. A lightweight text matching method with rich features is proposed for the problem of large number of model parameters and low efficiency of text matching tasks in natural language processing. The whole model architecture is based on Siamese neural networks with shared parameters. Furthermore, the method utilizes an improved residual Network and attention mechanism for the extraction and alignment of vector representations. Only three key features for alignment operations are retained. In addition, an averaging operation is added to the fusion layer to provide vector representations with rich information for the prediction layer. Experimental results on the paraphrase identification dataset and two natural language inference datasets show that the proposed approach not only effectively reduces the number of parameters compared with existing models but also ensures good text matching performance. Experiments demonstrate that this method can be used in general text matching tasks. © 2022 ACM.

关键词： Natural language processing systems

来源：评论

学校读者我要写书评

暂无评论

Plant-Leaf Disease Detection: A Comparative Analysis With Machine Learning and Deep Learning Methods

Plant-Leaf Disease Detection: A Comparative Analysis With Ma...

引用

Modeling, Simulation & Intelligent Computing (MoSICom), International conference on

作者： Varun Tiwari Amit Kumar Bairwa Shashi Bhushan Sharma Department of AIML Manipal University Jaipur Jaipur India Department of Computer Applications Manipal University Jaipur Jaipur India

ISBN: (数字)9798331533311

ISBN: (纸本)9798331533328

Agriculture is vital to human civilization, providing food and economic benefits. However, crops and plant leaves are vulnerable to various illnesses that hinder growth. Early and accurate identification of pathogens can prevent further damage, but current manual methods by farmers could be more efficient and precise, leading to potential crop destruction and reduced yields. Utilizing computerized image processing techniques can minimize losses and boost productivity. images of affected leaves or crops can be used to identify and classify plant diseases using a variety of techniques. Significant advancements have been made, particularly with deep learning (DL) and machine learning (ML) techniques, including convolutional neural networks (CNN), which are favored for their ability to process image information and understand spatial hierarchies. Selecting between traditional machine learning (ML) and deep learning (DL) approaches depends on the nature of the problem, the availability of data, and computational resources. For intricate tasks like image detection and classification, DL—particularly convolutional neural networks (CNNs)—is preferred when sufficient data and processing capabilities are accessible.

关键词： Deep learning Productivity Technological innovation Plant diseases Pathology Pathogens Crops Agriculture Convolutional neural networks Diseases

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500页 << < 491 492 493 494 495 496 497 498 499 500 > >>

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：