检索结果-内蒙古大学图书馆

Magnetic Resonance Imaging Reconstruction Based on Butterfly Dilated Geometric Distillation

Journal of Shanghai Jiaotong university (science) 2024年 1-10页

作者： Duo, Lin Xu, Boyu Ren, Yong Yang, Xin Faculty of Information Engineering and Automation Kunming University of Science and Technology Kunming650500 China

In order to improve the reconstruction accuracy of magnetic resonance imaging (MRI), an accurate natural image compressed sensing (CS) reconstruction network is proposed, which combines the advantages of model-based and deep learning-based CS-MRI methods. In theory, enhancing geometric texture details in linear reconstruction is possible. First, the optimization problem is decomposed into two problems: linear approximation and geometric compensation. Aimed at the problem of image linear approximation, the data consistency module is used to deal with it. Since the processing process will lose texture details, a neural network layer that explicitly combines image and frequency feature representation is proposed, which is named butterfly dilated geometric distillation network. The network introduces the idea of butterfly operation, skillfully integrates the features of image domain and frequency domain, and avoids the loss of texture details when extracting features in a single domain. Finally, a channel feature fusion module is designed by combining channel attention mechanism and dilated convolution. The attention of the channel makes the final output feature map focus on the more important part, thus improving the feature representation ability. The dilated convolution enlarges the receptive field, thereby obtaining more dense image feature data. The experimental results show that the peak signal-to-noise ratio of the network is 5.43 dB, 5.24 dB and 3.89 dB higher than that of ISTA-Net+, FISTA and DGDN networks on the brain data set with a Cartesian sampling mask CS ratio of 10%. © 2024, Shanghai Jiao Tong university.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

A laser powered EO sensor for measurement of EMP

引用

Optoelectronics Letters 2024年第12期20卷 709-713页

作者： ZHANG Jiahong MA Chao WANG Zhuo Faculty of Information Engineering and Automation Kunming University of Science and TechnologyKunming 650500China

This article presents an active electro-optic(EO)modulation electromagnetic pulse(EMP)sensor powered by *** sensor simulation model is established,and the temporal response characteristics of the sensor are analyzed based on finite element method(FEM).The laser powered supply circuit and receiving modulation circuit are designed and implemented.A monopole antenna is designed and used to revive the *** integrating the supply circuit,the receiving modulation circuit and the antenna together,the sensor is finally fabricated and *** results indicate that with the designed laser powered supply circuit,the sensor can operate continuously and *** measured standard 1.2/50μs lightning electromagnetic pulse(LEMP)in the time domain agrees well with the input LEMP voltage *** linear maximum and minimum measurable electric fields of the sensor are 20.8 kV/m and 221 V/m,*** the results demonstrate that the sensor can provide an effective technical means for the measurement of EMP in the time domain.

关键词： EMP sensor antenna

来源：评论

学校读者我要写书评

暂无评论

APST-Flow: A Reversible Network-Based Artistic Painting Style Transfer Method

引用

Computers, Materials & Continua 2023年第6期75卷 5229-5254页

作者： Meng Wang Yixuan Shao Haipeng Liu Faculty of Information Engineering and Automation Kunming University of Science and TechnologyKunming650500China

In recent years,deep generative models have been successfully applied to perform artistic painting style transfer(APST).The difficulties might lie in the loss of reconstructing spatial details and the inefficiency of model convergence caused by the irreversible en-decoder methodology of the existing *** to this,this paper proposes a Flow-based architecture with both the en-decoder sharing a reversible network *** proposed APST-Flow can efficiently reduce model uncertainty via a compact analysis-synthesis methodology,thereby the generalization performance and the convergence stability are *** the generator,a Flow-based network using Wavelet additive coupling(WAC)layers is implemented to extract multi-scale content ***,a style checker is used to enhance the global style consistency by minimizing the error between the reconstructed and the input *** enhance the generated salient details,a loss of adaptive stroke edge is applied in both the global and local model *** experimental results show that the proposed method improves PSNR by 5%,SSIM by 6.2%,and decreases Style Error by 29.4%over the existing models on the ChipPhi *** competitive results verify that APST-Flow achieves high-quality generation with less content deviation and enhanced generalization,thereby can be further applied to more APST scenes.

关键词： Artistic painting style transfer reversible network generative adversarial network wavelet transform

来源：评论

学校读者我要写书评

暂无评论

Element relational graph-augmented multi-granularity contextualized encoding for document-level event role filler extraction

引用

Frontiers of Computer science 2025年第2期19卷 137-138页

作者： Enchang ZHU Zhengtao YU Yuxin HUANG Shengxiang GAO Yantuan XIAN Faculty of Information Engineering and Automation Kunming University of Science and TechnologyKunming 650500China Yunnan Key Laboratory of Artificial Intelligence Kunming University of Science and TechnologyKunming 650500China

1 Introduction Document-level Role Filler Extraction aims to identify those spans of text that denote the role fillers for each event described in the document[1].Despite achieving certain accomplishments,existing methods are still not effective due to the following two issues:(1)there are difficulties in contextual modeling of long text,which requires modeling and understanding coherence and connections across sentences and paragraphs;(2)there usually ignore the explicit dependency relationships between event elements displayed in long *** this end,we propose a novel graph-augmented approach for document-level event role filler extraction,named element relational graph-augmented multi-granularity contextualized encoder(ERGM),whose main idea is to effectively enhance the model's capabilities in capturing deep semantic information of events in long texts and modeling dependency relationships among event elements by incorporating the Event elements relational ***,this method first constructs the structural graph by extracting elements from the source document.

关键词： contextual spite semantic

来源：评论

学校读者我要写书评

暂无评论

LKMT:Linguistics Knowledge-Driven Multi-Task Neural Machine Translation for Urdu and English

引用

Computers, Materials & Continua 2024年第10期81卷 951-969页

作者： Muhammad Naeem Ul Hassan Zhengtao Yu Jian Wang Ying Li Shengxiang Gao Shuwan Yang Cunli Mao Faculty of Information Engineering and Automation Kunming University of Science and TechnologyKunming650500China Yunnan Key Laboratory of Artificial Intelligence Kunming University of Science and TechnologyKunming650500China

Thanks to the strong representation capability of pre-trained language models,supervised machine translation models have achieved outstanding ***,the performances of these models drop sharply when the scale of the parallel training corpus is *** the pre-trained language model has a strong ability for monolingual representation,it is the key challenge for machine translation to construct the in-depth relationship between the source and target language by injecting the lexical and syntactic information into pre-trained language *** alleviate the dependence on the parallel corpus,we propose a Linguistics Knowledge-Driven MultiTask(LKMT)approach to inject part-of-speech and syntactic knowledge into pre-trained models,thus enhancing the machine translation *** the one hand,we integrate part-of-speech and dependency labels into the embedding layer and exploit large-scale monolingual corpus to update all parameters of pre-trained language models,thus ensuring the updated language model contains potential lexical and syntactic *** the other hand,we leverage an extra self-attention layer to explicitly inject linguistic knowledge into the pre-trained language model-enhanced machine translation *** on the benchmark dataset show that our proposed LKMT approach improves the Urdu-English translation accuracy by 1.97 points and the English-Urdu translation accuracy by 2.42 points,highlighting the effectiveness of our LKMT *** ablation experiments confirm the positive impact of part-of-speech and dependency parsing on machine translation.

关键词： Urdu NMT(neural machine translation) Urdu natural language processing Urdu Linguistic features low resources language linguistic features pretrain model

来源：评论

学校读者我要写书评

暂无评论

Machine Learning-Based Methods for Materials Inverse Design: A Review

引用

Computers, Materials & Continua 2025年第2期82卷 1463-1492页

作者： Yingli Liu Yuting Cui Haihe Zhou Sheng Lei Haibin Yuan Tao Shen Jiancheng Yin Faculty of Information Engineering and Automation Kunming University of Science and TechnologyKunming650093China Yunnan Key Laboratory of Computer Technologies Application Kunming University of Science and TechnologyKunming650500China Yunnan Tin Company Limited Tin Industry Branch Gejiu661000China Faculty of Materials Science and Engineering Kunming University of Science and TechnologyKunming650093China

Finding materials with specific properties is a hot topic in materials *** materials design relies on empirical and trial-and-error methods,requiring extensive experiments and time,resulting in high *** the development of physics,statistics,computer science,and other fields,machine learning offers opportunities for systematically discovering new *** through machine learning-based inverse design,machine learning algorithms analyze the mapping relationships between materials and their properties to find materials with desired *** paper first outlines the basic concepts of materials inverse design and the challenges faced by machine learning-based approaches to materials inverse ***,three main inverse design methods—exploration-based,model-based,and optimization-based—are analyzed in the context of different application ***,the applications of inverse design methods in alloys,optical materials,and acoustic materials are elaborated on,and the prospects for materials inverse design are *** authors hope to accelerate the discovery of new materials and provide new possibilities for advancing materials science and innovative design methods.

关键词： Materials inverse design machine learning target properties deep learning new materials discovery

来源：评论

学校读者我要写书评

暂无评论

Fast Parallel Magnetic Resonance Imaging Reconstruction Based on Sparsifying Transform Learning and Structured Low-Rank Model

引用

Journal of Shanghai Jiaotong university (science) 2023年 1-11页

作者： Duan, Jizhong Xu, Yuhan Huang, Huan Faculty of Information Engineering and Automation Kunming University of Science and Technology Kunming650500 China

The structured low-rank model for parallel magnetic resonance (MR) imaging can efficiently reconstruct MR images with limited auto-calibration signals. To improve the reconstruction quality of MR images, we integrate the joint sparsity and sparsifying transform learning (JTL) into the simultaneous auto-calibrating and k-space estimation (SAKE) structured low-rank model named JTLSAKE. The alternate direction method of multipliers is exploited to solve the resulting optimization problem, and the optimized gradient method is used to improve the convergence speed. In addition, a graphics processing unit is used to accelerate the proposed algorithm. The experimental results on four in vivo human datasets demonstrate that the reconstruction quality of the proposed algorithm is comparable to that of JTL-based low-rank modeling of local k-space neighborhoods with parallel imaging (JTL-PLORAKS), and the proposed algorithm is 46 times faster than the JTL-PLORAKS, requiring only 4 s to reconstruct a 200 × 200 pixels MR image with 8 channels. © 2023, Shanghai Jiao Tong university.

关键词： Magnetic resonance imaging

来源：评论

学校读者我要写书评

暂无评论

Improved Sensitivity Encoding Parallel Magnetic Resonance Imaging Reconstruction Algorithm Based on Efficient Sum of Outer Products Dictionary Learning

引用

Journal of Shanghai Jiaotong university (science) 2023年 1-11页

作者： Duan, Jizhong Su, Yan Faculty of Information Engineering and Automation Kunming University of Science and Technology Kunming650500 China

Sensitivity encoding (SENSE) is a parallel magnetic resonance imaging (MRI) reconstruction model by utilizing the sensitivity information of receiver coils to achieve image reconstruction. The existing SENSE-based reconstruction algorithms usually used nonadaptive sparsifying transforms, resulting in a limited reconstruction accuracy. Therefore, we proposed a new model for accurate parallel MRI reconstruction by combining the L 0 norm regularization term based on the efficient sum of outer products dictionary learning (SOUPDIL) with the SENSE model, called SOUPDIL-SENSE. The SOUPDIL-SENSE model is mainly solved by utilizing the variable splitting and alternating direction method of multipliers techniques. The experimental results on four human datasets show that the proposed algorithm effectively promotes the image sparsity, eliminates the noise and artifacts of the reconstructed images, and improves the reconstruction accuracy. © 2023, Shanghai Jiao Tong university.

关键词： Magnetic resonance imaging

来源：评论

学校读者我要写书评

暂无评论

Multispectral point cloud superpoint segmentation

引用

science China(Technological sciences) 2024年第4期67卷 1270-1281页

作者： WANG QingWang WANG MingYe ZHANG ZiFeng SONG Jian ZENG Kai SHEN Tao GU YanFeng Faculty of Information Engineering and Automation Kunming University of Science and TechnologyKunming 650500China Yunnan Key Laboratory of Computer Technologies Application Kunming 650500China School of Electronics and Information Engineering Harbin Institute of TechnologyHarbin 150001China

The multitude of airborne point clouds limits the point cloud processing *** are grouped based on similar points,which can effectively alleviate the demand for computing resources and improve processing ***,existing superpoint segmentation methods focus only on local geometric structures,resulting in inconsistent spectral features of points within a *** feature inconsistencies degrade the performance of subsequent ***,this study proposes a novel Superpoint Segmentation method that jointly utilizes spatial Geometric and Spectral information for multispectral point cloud superpoint segmentation(GSI-SS).Specifically,a similarity metric that combines spatial geometry and spectral information is proposed to facilitate the consistency of geometric structures and object attributes within segmented *** the formation of the primary superpoints,an intersuperpoint pointexchange mechanism that maximizes feature consistency within the final superpoints is *** are conducted on two real multispectral point cloud datasets,and the proposed method achieved higher recall,precision,F score,and lower global consistency and feature classification *** experimental results demonstrate the superiority of the proposed GSI-SS over several state-of-the-art methods.

关键词： multispectral point cloud superpoint segmentation over-segmentation spatial-spectral joint metric

来源：评论

学校读者我要写书评

暂无评论

Content and opinion-enhanced neural model for opinion sentence classification of Chinese microblog comments

引用

International Journal of information and Communication technology 2023年第4期23卷 371-387页

作者： Xiang, Yan Guo, Junjun Huang, Yuxin Yu, Zhengtao Faculty of Information Engineering and Automation Kunming University of Science and Technology Kunming650500 China

Opinion sentence classification of Chinese microblog comments aims to recognise those comments with opinions about the specific microblog content, which is the basis of internet public opinion analysis and opinion mining. However, existing opinion sentence classification methods do not consider whether the opinion sentences point at concerned objects or not. To address these issues, we propose a novel neural model, which combines the microblog content relevance-enhanced module and the opinion representation-enhanced module. In the first module, we propose a mutual attention operation that enables the model to extract better features representing microblog content. In the second module, we employ sentiment word embedding and self-attention operations to enhance the ability of the model to extract the opinion features. We evaluate our model using a Chinese microblog comment dataset. Experimental results show that the accuracy of the proposed model is 2%–5% higher than that of the baseline models, which shows that the proposed content relevance-enhancement and attention mechanism are beneficial to this task. Copyright © 2023 Inderscience Enterprises Ltd.

关键词： Sentiment analysis

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：