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

2,952 篇 会议
39 篇 期刊文献
7 册 图书

馆藏范围

2,998 篇 电子文献
0 种 纸本馆藏

日期分布

学科分类号

1,753 篇 工学
- 1,402 篇 计算机科学与技术...
- 683 篇 软件工程
- 488 篇 电气工程
- 394 篇 信息与通信工程
- 200 篇 控制科学与工程
- 130 篇 生物工程
- 78 篇 电子科学与技术（可...
- 65 篇 生物医学工程（可授...
- 58 篇 机械工程
- 54 篇 仪器科学与技术
- 48 篇 光学工程
- 33 篇 网络空间安全
- 30 篇 动力工程及工程热...
- 24 篇 安全科学与工程
- 23 篇 材料科学与工程（可...
- 22 篇 化学工程与技术
- 20 篇 测绘科学与技术
- 19 篇 石油与天然气工程
515 篇 理学
- 225 篇 物理学
- 188 篇 数学
- 143 篇 生物学
- 57 篇 统计学（可授理学、...
- 40 篇 系统科学
- 27 篇 化学
203 篇 管理学
- 135 篇 管理科学与工程(可...
- 81 篇 图书情报与档案管...
- 31 篇 工商管理
135 篇 医学
- 112 篇 临床医学
- 38 篇 基础医学(可授医学...
37 篇 法学
- 20 篇 社会学
18 篇 文学
14 篇 农学
10 篇 教育学
9 篇 经济学
2 篇 军事学

主题

590 篇 neural networks
173 篇 training
167 篇 artificial neura...
164 篇 deep learning
135 篇 computational mo...
124 篇 feature extracti...
115 篇 neural network
114 篇 convolutional ne...
88 篇 machine learning
87 篇 neurons
86 篇 parallel process...
79 篇 deep neural netw...
74 篇 computer archite...
73 篇 distributed comp...
65 篇 signal processin...
62 篇 distributed proc...
62 篇 accuracy
60 篇 recurrent neural...
55 篇 data models
49 篇 convolutional ne...

机构

11 篇 institute of inf...
9 篇 school of cyber ...
8 篇 univ chinese aca...
8 篇 natl univ def te...
7 篇 college of compu...
7 篇 carnegie mellon ...
7 篇 national laborat...
7 篇 beijing universi...
7 篇 natl univ def te...
6 篇 shanghai jiao to...
6 篇 the islamic univ...
6 篇 department of co...
5 篇 science and tech...
5 篇 ibm tj watson re...
5 篇 school of inform...
5 篇 king saud univer...
5 篇 shandong normal ...
5 篇 kaust thuwal
5 篇 university of ch...
5 篇 natl univ def te...

作者

8 篇 li dongsheng
8 篇 dou yong
8 篇 jie liu
6 篇 richtarik peter
6 篇 ribeiro alejandr...
6 篇 liu jie
6 篇 niu xin
6 篇 yong dou
6 篇 shulian yang
6 篇 zhang wei
5 篇 ryabinin max
5 篇 segarra santiago
5 篇 lai zhiquan
5 篇 ding bo
5 篇 qiao peng
5 篇 xu jiren
5 篇 kokkinos yiannis
5 篇 peng yuxing
5 篇 wang huaimin
5 篇 li aiping

语言

2,943 篇 英文
47 篇 其他
8 篇 中文
2 篇 土耳其文

检索条件"任意字段=Conference on Neural Network and Distributed Processing"

共 2998 条记录，以下是681-690 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：

Error Compensated distributed SGD can be Accelerated 35

Error Compensated Distributed SGD can be Accelerated

引用

35th Annual conference on neural Information processing Systems (NeurIPS)

作者： Qian, Xun Richtarik, Peter Zhang, Tong King Abdullah Univ Sci & Technol Thuwal Saudi Arabia Moscow Inst Phys & Technol Dolgoprudnyi Russia Hong Kong Univ Sci & Technol Hong Kong Peoples R China

ISBN: (纸本)9781713845393

Gradient compression is a recent and increasingly popular technique for reducing the communication cost in distributed training of large-scale machine learning models. In this work we focus on developing efficient distributed methods that can work for any compressor satisfying a certain contraction property, which includes both unbiased (after appropriate scaling) and biased compressors such as RandK and TopK. Applied naively, gradient compression introduces errors that either slow down convergence or lead to divergence. A popular technique designed to tackle this issue is error compensation/error feedback. Due to the difficulties associated with analyzing biased compressors, it is not known whether gradient compression with error compensation can be combined with acceleration. In this work, we show for the first time that error compensated gradient compression methods can be accelerated. In particular, we propose and study the error compensated loopless Katyusha method, and establish an accelerated linear convergence rate under standard assumptions. We show through numerical experiments that the proposed method converges with substantially fewer communication rounds than previous error compensated algorithms.

关键词： Error compensation

来源：评论

学校读者我要写书评

暂无评论

Real-Time Audio Noise Reduction and Speech Enhancement Using LadderNet With Hybrid Spectrogram Time-Domain Audio Separation network

Real-Time Audio Noise Reduction and Speech Enhancement Using...

引用

Mobile networks and Wireless Communications (ICMNWC), IEEE International conference on

作者： Ayesha Siddiqua CH Hussaian Basha Haider Mohammed Abbas P. Merlin Sheela V. Indhumathi Department of Electronics and Communication Engineering Nitte Meenakshi Institute of Technology Bengaluru India Department of Electrical & Electronics Engineering School of Engineering SR University Warangal India Department of Computers Techniques Engineering College of Technical Engineering The Islamic University Najaf Iraq Department of English New Prince Shri Bhavani College of Engineering and Technology Chennai India Department of Computer Applications Sona College of Arts and Science Salem India

ISBN: (数字)9798331528348

ISBN: (纸本)9798331528355

In recent years, the rapid growth of modern technology has improved people lives, but it also comes with a downside: increased exposure to noise from complex industrial infrastructure and machinery. To overcome this issue, the existing method namely a single-channel Deep neural network (DNN) model have developed for speech signal processing, but it led to limitations in online processing. Therefore, this research proposes a LadderNet with Hybrid Spectrogram Time-domain Audio Separation network (HSTasNet) model for reducing the noise and enhancing speech enhancement in real-time audio processing. This procedure begins with the collection of data from real-time audio signal and then preprocessed with spectral gating filter to reduce noise from signals. After that, t-distributed Stochastic Neighbor Embedding (t-SNE) technique to extract the optimal features from preprocessed signals. Finally, LadderNet and HS-TasNet techniques are employed for noise reduction and speech enhancement. As per the results, the proposed LadderNet-HSTasNet model offered outstanding results in terms of Perceptual Evaluation of Speech Quality (PESQ) of 3.847, Short-Time Objective Intelligibility (STOI) of 1.874, and Scale-Invariant Signal-to-Distortion Ratio (SI-SDR) of 16.792 when compared with existing Dual-Path Recurrent neural network (DPRNN) model.

关键词： Wireless communication Recurrent neural networks Noise reduction Stochastic processes Artificial neural networks Speech enhancement Feature extraction Real-time systems Time-domain analysis Spectrogram

来源：评论

学校读者我要写书评

暂无评论

Deep neural network Architectures for Spectrum Sensing Using Signal processing Features

Deep Neural Network Architectures for Spectrum Sensing Using...

引用

2021 IEEE International conference on distributed Computing, VLSI, Electrical Circuits and Robotics, DISCOVER 2021

作者： Chandra, Shreeram Suresh Upadhye, Akshay Saravanan, Purushothaman Gurugopinath, Sanjeev Muralishankar, R. PES University Department of Electronics and Communication Engineering Bengaluru560085 India CMR University School of Engineering and Technology Bengaluru562149 India

ISBN: (纸本)9781665412445

In this work, we consider a performance comparison of deep learning-based approaches to the problem of spectrum sensing (SS) in cognitive radios. Towards this end, we use signal processing (SP) features such as energy, differential entropy, geometric power and p-norm. For the classification problem of SS, we employ deep neural network (NN) architectures such as multi-layer perceptron (MLP), convolutional NN, fully convolutional network, residual NN (ResNet), long short-Term memory and temporal convolutional network. Through extensive experiments based on real-world captured datasets and Monte Carlo simulations, we show that MLP and ResNet architectures offer the best performance in terms of probability of detection, for a given predefined level of probability of false-Alarm. Further, we show that NN architectures trained with a combined set of the SP features yield the best performance. © 2021 IEEE.

关键词： Multilayer neural networks

来源：评论

学校读者我要写书评

暂无评论

A MULTI-TASK CONVOLUTIONAL neural network FOR BLIND STEREOSCOPIC IMAGE QUALITY ASSESSMENT USING NATURALNESS ANALYSIS

A MULTI-TASK CONVOLUTIONAL NEURAL NETWORK FOR BLIND STEREOSC...

引用

IEEE International conference on Image processing (ICIP)

作者： Bourbia, Salima Karine, Ayoub Chetouani, Aladine El Hassoun, Mohammed Mohammed V Univ Rabat LRIT Rabat Morocco ISEN Yncrea Ouest L bISEN 33 Quater Chemin Champ Manoeuvre F-44470 Carquefou France Univ Orleans Lab PRISME Orleans France Mohammed V Univ Rabat FLSH Rabat Morocco

ISBN: (纸本)9781665441155

This paper addresses the problem of blind stereoscopic image quality assessment (NR-SIQA) using a new multi-task deep learning based-method. In the field of stereoscopic vision, the information is fairly distributed between the left and right views as well as the binocular phenomenon. In this work, we propose to integrate these characteristics to estimate the quality of stereoscopic images without reference through a convolutional neural network. Our method is based on two main tasks: the first task predicts naturalness analysis based features adapted to stereo images, while the second task predicts the quality of such images. The former, so-called auxiliary task, aims to find more robust and relevant features to improve the quality prediction. To do this, we compute naturalness-based features using a Natural Scene Statistics (NSS) model in the complex wavelet domain. It allows to capture the statistical dependency between pairs of the stereoscopic images. Experiments are conducted on the well known LIVE PHASE I and LIVE PHASE II databases. The results obtained show the relevance of our method when comparing with those of the state-of-the-art.

关键词： Blind stereoscopic image quality assessment Convolutional neural network Multi-task deep learning Naturalness based-features Binocular features

来源：评论

学校读者我要写书评

暂无评论

Convolutional Block Attention Mechanism with ResNet18 for Sports Action Recognition based on Video Image Key Point Detection

Convolutional Block Attention Mechanism with ResNet18 for Sp...

引用

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

作者： Chuanming Xu Lu Sun School of Arts and Sciences Chengdu College of University of Electronic Science and Technology of China Chengdu China School of Computer Science Chengdu College of University of Electronic Science and Technology of China Chengdu China

ISBN: (数字)9798350375442

ISBN: (纸本)9798350375459

Sports action recognition and evaluation depending on video image key point detection contains analyzing athletes’ movements by determining critical body joint and landmark in video frames. These key points are utilized to track motion and enable accurate classification and evaluation of different sports actions. However, it is challenging due to variability in body movements, diverse action style, and occlusions across athletes. This research proposes Convolutional Block Attention Mechanism-ResNet18 to recognize and evaluate the sports action depending on video image key point detection. In ResNet18, the CBAM is included to enhance feature representation by focusing on significant spatial and channel-wise data which improves the network’s ability to capture appropriate data. Initially, the UCF-101 dataset is used to determine the proposed approach performance. The mean subtraction is applied in pre-processing to normalize input data by centering it around zero which enhances model convergence and training stability. The proposed CBAM-ResNet18 achieves a better accuracy of 98.73% using UCF-101 dataset when compared to existing methods like Convolutional neural network (CNN) and Bidirectional-Gated Recurrent Unit (Bi-GRU) and Attention-based two-stream deep neural network (A2SN).

关键词： Training Accuracy Image recognition Attention mechanisms Tracking Stability analysis Spatial databases Convolutional neural networks Sports Convergence

来源：评论

学校读者我要写书评

暂无评论

Towards Tight Communication Lower Bounds for distributed Optimisation 35

Towards Tight Communication Lower Bounds for Distributed Opt...

引用

35th Annual conference on neural Information processing Systems (NeurIPS)

作者： Alistarh, Dan Korhonen, Janne H. IST Austria Klosterneuburg Austria Neural Mag Inc Somerville MA 02144 USA

ISBN: (纸本)9781713845393

We consider a standard distributed optimisation setting where N machines, each holding a d-dimensional function f(i), aim to jointly minimise the sum of the functions Sigma(N)(i=1) f(i)(x). This problem arises naturally in large-scale distributed optimisation, where a standard solution is to apply variants of (stochastic) gradient descent. We focus on the communication complexity of this problem: our main result provides the first fully unconditional bounds on total number of bits which need to be sent and received by the N machines to solve this problem under point-to-point communication, within a given error-tolerance. Specifically, we show that Omega(N d log d/N epsilon) total bits need to be communicated between the machines to find an additive epsilon-approximation to the minimum of Sigma(N)(i=1) f(i)(x). The result holds for both deterministic and randomised algorithms, and, importantly, requires no assumptions on the algorithm structure. The lower bound is tight under certain restrictions on parameter values, and is matched within constant factors for quadratic objectives by a new variant of quantised gradient descent, which we describe and analyse. Our results bring over tools from communication complexity to distributed optimisation, which has potential for further applications.

关键词： Computational complexity

来源：评论

学校读者我要写书评

暂无评论

PS-Based Heterogeneous Computing Framework

PS-Based Heterogeneous Computing Framework

引用

Next Generation Data-driven networks (NGDN), International conference on

作者： Xi Guo Ming He Shouhao Qin Dexuan Wang Chenjun Li Research Institute China United Network Communications Corporation Beijing China GuoChuang Software Co. Ltd. Hefei China

ISBN: (数字)9798350388374

ISBN: (纸本)9798350388381

In the current heterogeneous computing environment, different types of accelerated computing resources such as GPU(Graphics processing Unit) and NPU(neural processing Unit) coexist in the cluster. However, because these resources use different communication protocols and architectures, heterogeneous computing resources cannot effectively communicate and collaborate. To solve this problem, we propose a heterogeneous distributed computing framework based on the Parameter Server(PS) architecture and achieve heterogeneous parallel computing capabilities across hardware devices. This allows users to uniformly use all available heterogeneous computing resources in the cluster to perform large-scale data parallel deep learning training tasks. We have conducted experiments and evaluations on this framework in multiple scenarios, and the results show that the NPU and GPU computing power in a heterogeneous computing cluster can be used simultaneously to complete a same computing task without affecting accuracy, while ensuring high training efficiency. Overall, this framework provides an effective solution for high-performance distributed computing in heterogeneous computing environments.

关键词： Training Deep learning Performance evaluation Protocols Processor scheduling Graphics processing units Computer architecture Heterogeneous networks distributed computing Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

Dynamic DNN Decomposition for Lossless Synergistic Inference 41

Dynamic DNN Decomposition for Lossless Synergistic Inference

引用

IEEE 41st International conference on distributed Computing Systems (ICDCS)

作者： Zhang, Beibei Xiang, Tian Zhang, Hongxuan Li, Te Zhu, Shigiang Gu, Jianjun Zhejiang Lab Intelligent Robot Res Ctr Hangzhou Peoples R China

ISBN: (纸本)9781665449328

Deep neural networks (DNNs) sustain high performance in today's data processing applications. DNN inference is resource-intensive thus is difficult to fit into a mobile device. To solve this problem, recent advances unleash DNN services using the edge computing paradigm. The existing approaches split a DNN into two parts and deploy the two partitions to computation nodes at two edge computing tiers. Nonetheless, these methods overlook collaborative device-edge-cloud computation resources. Besides, previous algorithms demand the whole DNN re-partitioning to adapt to computation resource changes and network dynamics. Moreover, for resource-demanding convolutional layers, prior works do not give a parallel processing strategy without loss of accuracy at the edge side. To tackle these issues, we propose D-3, a dynamic DNN decomposition system for synergistic inference without precision loss. The proposed system introduces a heuristic algorithm named horizontal partition algorithm to split a DNN into three parts. The algorithm partially adjust the partitions at run time according to processing time and network conditions. At the edge side, a vertical separation module separates feature maps into tiles that can be independently run on different edge nodes in parallel. Extensive quantitative evaluation of five popular DNNs illustrates that D-3 outperforms the stateof-the-art counterparts up to 3.4X in end-to-end DNN inference time and reduces backbone network communication overhead up to 3.68X.

关键词： distributed computing Edge computing DNN inference acceleration

来源：评论

学校读者我要写书评

暂无评论

ENIDrift: A Fast and Adaptive Ensemble System for network Intrusion Detection under Real-world Drift 22

ENIDrift: A Fast and Adaptive Ensemble System for Network In...

引用

38th Annual Computer Security Applications conference (ACSAC)

作者： Wang, Xian Hong Kong Univ Sci & Technol Hong Kong Peoples R China

ISBN: (纸本)9781450397599

Machine Learning (ML) techniques have been widely applied for network intrusion detection. However, existing ML-based network intrusion detection systems (NIDSs) suffer from fundamental limitations that hinder them from being deployed in the real world. They consider a narrow scope rather than real-world drift that involves dynamically distributed network packets and well-crafted ML attacks. Besides, they pose high runtime overhead and have low processing speed. In this paper, we solve the limitations and design ENIDrift, a fast and adaptive ensemble system for real-world network intrusion detection. ENIDrift employs iP2V, a novel incremental feature extraction method based on network packet fields, which adopts a simple three-layer neural network with relatively lightweight computation and achieves high efficiency. ENIDrift uses a stable sub-classifier generation module that constructs new sub-classifiers based on the stability and accuracy of incoming data chunks, and its training time is reduced from O(..) to O(1). We extend the threat model and place experiments in real-world settings. We make the first real-world drift dataset, RWDIDS, which contains intense and different drifts for NIDS. Our extensive evaluation under real-world drift demonstrates that ENIDrift significantly outperforms the stateof-the-art solutions by up to 69.78% of F1 and reduces running time by 87.6%. ENIDrift achieves a 100% F1 against our adversarial attack and is adaptive to various real-world drifts. Our field test also shows ENIDrift functionswell even with delayed, inadequate training data, which is practical for real-world usage.

关键词： network Intrusion Detection network Security Anomaly Detection Ensemble Learning Concept Drift

来源：评论

学校读者我要写书评

暂无评论

neural UTTERANCE CONFIDENCE MEASURE FOR RNN-TRANSDUCERS AND TWO PASS MODELS

NEURAL UTTERANCE CONFIDENCE MEASURE FOR RNN-TRANSDUCERS AND ...

引用

IEEE International conference on Acoustics, Speech and Signal processing (ICASSP)

作者： Gupta, Ashutosh Gowda, Ankur Kumar Dhananjaya Kim, Kwangyoun Singh, Sachin Singh, Shatrughan Kim, Chanwoo Samsung Res Seoul South Korea

ISBN: (纸本)9781728176055

In this paper, we propose methods to compute confidence score on the predictions made by an end-to-end speech recognition model in a 2-pass framework. We use RNN-Transducer for a streaming model, and an attention-based decoder for the second pass model. We use neural technique to compute the confidence score, and experiment with various combinations of features from RNN-Transducer and second pass models. The neural confidence score model is trained as a binary classification task to accept or reject a prediction made by speech recognition model. The model is evaluated in a distributed speech recognition environment, and performs significantly better when features from second pass model are used as compared to the features from streaming model.

关键词： neural confidence measure end-to-end speech recognition RNN-Transducers Two pass

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300页 << < 65 66 67 68 69 70 71 72 73 74 > >>

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：