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

154 篇 会议
44 篇 期刊文献

馆藏范围

198 篇 电子文献
0 种 纸本馆藏

日期分布

学科分类号

118 篇 工学
- 79 篇 计算机科学与技术...
- 74 篇 软件工程
- 51 篇 信息与通信工程
- 21 篇 生物工程
- 20 篇 光学工程
- 14 篇 化学工程与技术
- 12 篇 生物医学工程（可授...
- 9 篇 仪器科学与技术
- 8 篇 电气工程
- 8 篇 电子科学与技术（可...
- 8 篇 控制科学与工程
- 7 篇 机械工程
- 6 篇 安全科学与工程
- 4 篇 材料科学与工程（可...
- 4 篇 土木工程
- 3 篇 动力工程及工程热...
- 3 篇 建筑学
- 3 篇 环境科学与工程（可...
74 篇 理学
- 33 篇 物理学
- 26 篇 数学
- 23 篇 生物学
- 16 篇 化学
- 9 篇 统计学（可授理学、...
- 4 篇 大气科学
36 篇 管理学
- 27 篇 图书情报与档案管...
- 9 篇 管理科学与工程(可...
8 篇 医学
- 8 篇 基础医学(可授医学...
- 8 篇 临床医学
- 7 篇 药学(可授医学、理...
4 篇 教育学
- 4 篇 教育学

主题

26 篇 feature extracti...
12 篇 deep learning
12 篇 training
10 篇 neural networks
10 篇 convolution
10 篇 accuracy
9 篇 speech processin...
9 篇 semantics
6 篇 speech enhanceme...
6 篇 data models
6 篇 adaptation model...
5 篇 object detection
5 篇 noise reduction
5 篇 transformers
5 篇 noise
5 篇 attention mechan...
4 篇 raman spectrosco...
4 篇 noise measuremen...
4 篇 signal processin...
4 篇 acoustics

机构

54 篇 school of comput...
43 篇 xinjiang key lab...
27 篇 key laboratory o...
16 篇 school of inform...
13 篇 department of el...
10 篇 xinjiang key lab...
9 篇 school of intell...
8 篇 school of inform...
7 篇 department of el...
7 篇 key laboratory o...
7 篇 key laboratory o...
7 篇 china mobile res...
7 篇 college of infor...
6 篇 xinjiang key lab...
6 篇 college of softw...
6 篇 college of softw...
5 篇 school of comput...
5 篇 key laboratory o...
5 篇 xinjiang key lab...
4 篇 college of compu...

作者

23 篇 hamdulla askar
22 篇 jia zhenhong
12 篇 liang he
12 篇 shi fei
12 篇 he liang
12 篇 askar hamdulla
11 篇 gang zhou
11 篇 zhenhong jia
10 篇 wang jiajia
10 篇 zhou gang
9 篇 ying hu
9 篇 chen chen
9 篇 lv xiaoyi
9 篇 qin xizhong
9 篇 zhihua huang
8 篇 chen cheng
7 篇 fei shi
7 篇 huang xiaohui
7 篇 hu ying
7 篇 ablimit mijit

语言

194 篇 英文
3 篇 其他
2 篇 中文
1 篇 德文
1 篇 法文
1 篇 意大利文

检索条件"机构=Key Laboratory of Signal Detection and Processing"

共 198 条记录，以下是111-120 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：

Improved Self-Consistency Training with Selective Feature Fusion for Sound Event detection

Improved Self-Consistency Training with Selective Feature Fu...

引用

IEEE International Conference on Information Communication and signal processing (ICICSP)

作者： Mingyu Wang Yunlong Li Ying Hu Key Laboratory of signal detection and processing Xinjiang University Urumqi China

Sound event detection (SED) is a joint task of identifying the categories and time boundaries of sound events within an audio clip. In this paper, we propose an improved self-consistency training (ISCT) strategy for semi-supervised SED based on Mean Teacher (MT) method. For teacher and student models, each adopts two branches with the same CRNN structure, the two branches help training the model by means of consistency regularization. ISCT strategy incorporates self-consistency loss on the basis of MT loss to improve the generalization performance of the model. A selective feature fusion (SFF) module is designed for applying in the shallow layers of the feature extraction part to selectively fuse the features with different scales. A parallel attention (PA) module is designed for applying in the deep layers of the feature extraction part to obtain much richer high-level features by the channel and spatial-wise attention. Ablation experiments verify the effectiveness of our proposed ISCT strategy, SFF and PA modules. In addition, compared with four methods, our proposed method achieves competitive performance on the DCASE 2020 task4 dataset.

关键词：

来源：评论

学校读者我要写书评

暂无评论

EEMD and Double Thresholds Integrated Voice Activity detection 3

EEMD and Double Thresholds Integrated Voice Activity Detecti...

引用

3rd International Conference on Pattern Recognition and Machine Learning, PRML 2022

作者： Meng, Shan Ablimit, Mijit Hamdulla, Askar Xinjiang University Xinjiang Key Laboratory of Signal Detection and Processing School of Information Science and Engineering Xinjiang Urumqi China Xinjiang University Xinjiang Key Laboratory of Multilingual Information Technology School of Information Science and Engineering Xinjiang Urumqi China

ISBN: (数字)9781665499507

ISBN: (纸本)9781665499507

Voice activity detection (VAD) is an important preprocessing for voice applications. Anti-noise performance is the most important evaluation index of VAD algorithm. The traditional dual-threshold-based VAD algorithm has very low detection accuracy in a low signal-to-noise ratio environment. This paper proposes a voice activity detection algorithm based on Ensemble Empirical Mode Decomposition (EEMD) combined with the dual-threshold method, which integrates the decomposition of EEMD. The denoising feature is combined with VAD based on dual thresholds, and dual thresholds are set for VAD to improve the anti-noise performance and accuracy of the algorithm. VAD is divided into three categories: based on feature parameters, based on pattern recognition, and based on deep learning. The VAD algorithms based on pattern recognition and deep learning require the support of big training data to achieve good detection results. And it is too complex and requires a large amount of computation, thus its application and real-time deployment have been greatly affected. The traditional VAD methods based on feature parameters only needs to use the short-term energy and short-term zero-crossing rate as the judgment criteria for activity detection. Thise algorithm is simple, easy to deploy, and applicable even if it is dozens of voices. At the same time, the EEMD changes the extreme point characteristics of the signal by adding different white noises of the same amplitude each time, and then performs the overall average of the corresponding IMF(intrinsic mode function) obtained by multiple EMDs to offset the added white noise, therefor effectively suppressing the mode-mixing. Thus the production of state aliasing can better improve the anti-noise performance of the VAD algorithm. © 2022 IEEE.

关键词： Empirical mode decomposition

来源：评论

学校读者我要写书评

暂无评论

Research of Scene Text detection Algorithms 5

Research of Scene Text Detection Algorithms

引用

5th International Conference on Intelligent Robotics and Control Engineering, IRCE 2022

作者： Chen, Mengmeng Ibrayim, Mayire Hamdulla, Askar School of Information Science and Engineering Xinjiang University Xinjiang Key Laboratory of Signal Detection and Processing Xinjiang Urumqi China School of Information Science and Engineering Xinjiang University Xinjiang Key Laboratory of Multilingual Information Technology Xinjiang Urumqi China

ISBN: (数字)9781665469951

ISBN: (纸本)9781665469951

With the development of artificial intelligence, obtaining textual information from natural scenes has become a hot topic. There are still huge challenges for curved text and arbitrary orientation text detection in real-world scenes. In this work, we briefly introduce the development of text detection algorithms in recent years and study a typical lightweight detection algorithm. We have improved the algorithm to different degrees to make it more suitable for detection of the curved text. Finally, we select the three most representative public text datasets for experiments. Experiments conducted on CTW1500, ICDAR2015, and MSRA-TD500 have demonstrated the superiority of this model. © 2022 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Residualpath-res-dense-net for retinal vessel segmentation 2

Residualpath-res-dense-net for retinal vessel segmentation

引用

2nd International Conference on Computer Vision, Image, and Deep Learning

作者： Huang, Xiaoming He, Fuyun Tang, Xiaohu Wang, Xun Qiu, Senhui Hu, Cong College of Electronic Engineering Guangxi Normal University Guilin541004 China Guangxi Key Laboratory of Automatic Detection Technology and Instrument Guilin541004 China Guangxi Key Laboratory of Wireless Wideband Communication and Signal Processing Guilin541004 China

ISBN: (数字)9781510646827

ISBN: (纸本)9781510646810

Retinal blood vessels segmentation acts as an important part to the treatment of ocular disease. Lately, automatic segmentation based on deep learning can solves problems of low efficiency and strong subjectivity of manual segmentation, and attracts the attention of researchers. In this paper, a novel segmentation model called Residualpath-Res-Dense Net (RRD-Net) is proposed to achieve vessel segmentation. In RRD-Net, Res-block and Dense-block are used to help speeding up the convergence of the network and learning more intrinsic features. In addition, the introduction of Residual Path can cut down the semantic difference between the connected feature and eliminate the potential impact of semantic difference on segmentation accuracy. We apply benchmark datasets DRIVE and CHASE_DB1 to evaluate effectiveness of the proposed network. Accuracy, sensitivity and F1 score demonstrate the effectiveness of RRD-Net. © 2021 SPIE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

RoPE-BAM: Nested Entity Recognition Based on Rotary Position Embedding and Biaffine Attention Mechanism 3

RoPE-BAM: Nested Entity Recognition Based on Rotary Position...

引用

3rd International Conference on Advanced Algorithms and Neural Networks, AANN 2023

作者： Deng, Jinxin Qin, Xizhong Yang, Rong Lv, Xiaoyi School of Information Science and Engineering Xinjiang University of China Xinjiang Key Laboratory of Signal Detection and Processing Urumqi830017 China The Fifth Affiliated Hospital of Xinjiang Medical Urumqi830017 China

ISBN: (纸本)9781510668355

Named entity recognition (NER) involves two main types: nested NER and flat NER. The span-based approach classifies entity types by head-tail pair span representations and can handle nested and flat entities uniformly. However, the span-based approach uses a single feature and ignores the relative position information between head-tail pairs, which affects the precision of entity recognition. Therefore, we propose a nested entity recognition model that combines rotary position embedding and biaffine attention mechanism (RoPE-BAM) to improve the model performance by adding relative position features to the span representations. Concretely, we first obtain the head sequence and tail sequence representations through two feedforward networks. Then, to incorporate the relative position features, rotary position embedding is applied to both head and tail sequences. Finally, we use a biaffine attention mechanism to capture the span representations while generating the relative position information in the span. Extensive experiments were conducted on five widely-used benchmark datasets to demonstrate the effectiveness of our proposed RoPE-BAM model. © 2023 SPIE.

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

AMIMC: Based on Improved Meta-Learning Contrast Network with Attention Mechanism for Few Shot Text Classification

AMIMC: Based on Improved Meta-Learning Contrast Network with...

引用

Cloud Computing, Big Data Applications and Software Engineering (CBASE), International Conference on

作者： Senyan Li Ye Dong Xizhong Qin Xinjiang Signal Detection and Processing Key Laboratory Xinjiang University Urumqi China

ISBN: (数字)9798331506582

ISBN: (纸本)9798331506599

Currently, meta-learning is the mainstream approach to solving the problem of scarce data in few-shot text classification. Still, challenges remain, such as embedding vectors not being compact enough, suboptimal meta-task sampling strategies, and the rigidity of convolution operations in traditional Text CNNs. To address this, we propose the Attention Mechanism-based Improved Meta-learning Contrast Network (AMIMC), which enhances intra-class aggregation, increases inter-class separation and improves embedding quality. Additionally, Double Dynamic Similar Sampling (DDSS) generates more challenging meta-tasks, and the attention mechanism enhances the flexibility of Text CNNs, significantly boosting accuracy on five few-shot text classification datasets.

关键词： Metalearning Attention mechanisms Reviews Text categorization Semantics Vectors Robustness Rigidity Noise measurement Software engineering

来源：评论

学校读者我要写书评

暂无评论

MetricGAN+ Speech Enhancement Model Combining Multi-scale and Multi-resolution Feature Network and Joint Perception Loss

MetricGAN+ Speech Enhancement Model Combining Multi-scale an...

引用

Pattern Recognition and Machine Learning (PRML), IEEE International Conference on

作者： Chuangjian Guo Zhihua Huang Hui Li Jingyi Song Key Laboratory of Signal Detection and Processing in Xinjiang Uygur Autonomous Region Xinjiang University Urumchi China

Recently, generative adversarial networks have been widely used in speech enhancement. However, some existing speech enhancement methods only aim to optimize a single perceptual metric, which may cause other perceptual metrics to be suboptimal. In this paper, we propose a joint perceptual loss function to guide the training of speech enhancement models. In addition, we use a multi-scale and multi-resolution feature extraction network in the discriminator of MetriGAN+model, so that the model can make full use of high-level context information and low-level time-frequency features, so as to improve the prediction performance of the discriminator and guide the training of the generator better. The evaluation results on the VoiceBank-DEMAND dataset show that the proposed model can effectively improve speech quality and intelligibility, and can improve the PESQ score by 0.19 compared to the previous MetricGAN+.

关键词：

来源：评论

学校读者我要写书评

暂无评论

RFBiCF: A Relation-First Bidirectional Cascade Framework for Relational Triple Extraction 4

RFBiCF: A Relation-First Bidirectional Cascade Framework for...

引用

4th IEEE International Conference on Pattern Recognition and Machine Learning, PRML 2023

作者： Jinxin, Deng Xizhong, Qin Rong, Yang Xiaoyi, Lv Xinjiang University of China School of Information Science and Engineering Urumqi830017 China Xinjiang Key Laboratory of Signal Detection and Processing Urumqi830017 China School of Information the Fifth Affiliated Hospital of Xinjiang Medical Urumqi830017 China

ISBN: (纸本)9798350324303

Relational triple extraction from unstructured text, which is in form of (subject, relation, object), is crucial in information extraction and knowledge graph construction. However, most existing methods still suffer from some limitations, such as error propagation, redundancy of relation prediction, and poor generalization. To overcome these deficiencies, we propose a relation-first bidirectional cascade multi-Task framework for relational triple extraction. This framework solves the relational redundancy problem by relation prediction, alleviates error propagation through a complementary bidirectional entity tagger, and uses multi-Task learning to improve the generalization of the model. Concretely, we first use a relation detection component to predict potential relations in sentences and conduct the following entity pairs extraction based on specific relations. Then, we extract entity pairs from two paralleled directions. Meanwhile, entity recognition is used as an auxiliary task, improving generalization capability. Extensive experiments were conducted on the NYT and WebNLG datasets We evaluate our model on four public datasets to demonstrate the effectiveness of our proposed RFBiCF framework, even in the complex scenario of overlapping triples. © 2023 IEEE.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

keyword Spotting Based on Efficient Neural Architecture Search

Keyword Spotting Based on Efficient Neural Architecture Sear...

引用

Pattern Recognition and Machine Learning (PRML), IEEE International Conference on

作者： Wenchen Liu Zhihua Huang Dafei Wang Key Laboratory of Signal Detection and Processing in Xinjiang Uygur Autonomous Region Xinjiang University Urumchi China

In this paper, neural architecture search (NAS) technique is used to develop an efficient and low consumption keyword spotting model on Uyghur language dataset. The model obtained in this paper using NAS on the Google speech command dataset has advantages compared with other models. The model obtained by applying the ProxylessNAS technique to optimize the neural architecture search achieves more than 95% accuracy on the Uyghur dataset with less than 400 Kb memory *** factors affecting the model performance are also explored in this *** experimental results show that decreasing the model weights can significantly reduce the model memory, and the accuracy rate decreases slightly. Meanwhile, increasing the feature dimension of the input MFCC can improve the model accuracy, and the accuracy can reach 96.80% at 40 MFCC.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Fast Sampling Based on Policy Gradient for Diffusion-Based Speech Enhancement

Fast Sampling Based on Policy Gradient for Diffusion-Based S...

引用

International Symposium on Chinese Spoken Language processing

作者： Yubo Jiang Zhihua Huang School of Computer Science and Technology Xinjiang University Xinjiang Key Laboratory of Signal Detection and Processing Urumqi

ISBN: (数字)9798331516826

ISBN: (纸本)9798331516833

The effectiveness of diffusion-based generative models in speech enhancement tasks has been reported. These generative methods can produce high-quality audio with less distortions and outperform their discriminative counterparts, especially in mismatched conditions. However, due to the multi-step denoising property, these models usually have slower audio generation speed and a tendency to accumulate errors in the enhancement process. To address these issues, we propose a novel approach that considers the reverse process as a sequential decision-making task and constructs the distribution similarity metric into the training objective. In this way, the generator can learn data distributions well and explore more efficient sampling trajectories. The results of our experiments indicate that our proposed method not only reduces the number of steps required for audio generation, but it also produces even higher quality samples and preserves the generalizability to unknown noise.

关键词： Training Measurement Noise reduction Noise Reinforcement learning Speech enhancement Probabilistic logic Distortion Generators Trajectory

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20页 << < 8 9 10 11 12 13 14 15 16 17 > >>

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：