检索结果-南通市图书馆

A deformable Network with Attention Mechanism for Retinal Vessel Segmentation

Journal of Beijing Institute of Technology 2024年第3期33卷 186-193页

作者： Xiaolong Zhu Wenjian Li Weihang Zhang Dongwei Li Huiqi Li School of Information and Electronics Beijing Institute of TechnologyBeijing 100081China

The intensive application of deep learning in medical image processing has facilitated the advancement of automatic retinal vessel segmentation *** overcome the limitation that traditional U-shaped vessel segmentation networks fail to extract features in fundus image sufficiently,we propose a novel network(DSeU-net)based on deformable convolution and squeeze excitation residual *** deformable convolution is utilized to dynamically adjust the receptive field for the feature extraction of retinal *** the squeeze excitation residual module is used to scale the weights of the low-level features so that the network learns the complex relationships of the different feature layers *** validate the DSeU-net on three public retinal vessel segmentation datasets including DRIVE,CHASEDB1,and STARE,and the experimental results demonstrate the satisfactory segmentation performance of the network.

关键词： retinal vessel segmentation deformable convolution attention mechanism deep learning

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

DT-Net:Joint Dual-Input Transformer and CNN for Retinal Vessel Segmentation

引用

Computers, Materials & Continua 2023年第9期76卷 3393-3411页

作者： Wenran Jia Simin Ma Peng Geng Yan Sun School of Information Science and Technology Shijiazhuang Tiedao UniversityShijiazhuangChina School of Mathematics and Information Science Zhangjiakou UniversityZhangjiakouChina

Retinal vessel segmentation in fundus images plays an essential role in the screening,diagnosis,and treatment of many *** acquired fundus images generally have the following problems:uneven illumination,high noise,and complex *** makes vessel segmentation very *** methods of retinal vascular segmentation mainly use convolutional neural networks on U Network(U-Net)models,and they have many limitations and shortcomings,such as the loss of microvascular details at the end of the *** address the limitations of convolution by introducing the transformer into retinal vessel ***,we propose a hybrid method for retinal vessel segmentation based on modulated deformable convolution and the transformer,named ***,multi-scale image features are extracted by deformable convolution and multi-head selfattention(MHSA).Secondly,image information is recovered,and vessel morphology is refined by the proposed transformer decoder ***,the local prediction results are obtained by the side output *** accuracy of the vessel segmentation is improved by the hybrid loss *** results show that our method obtains good segmentation performance on Specificity(SP),Sensitivity(SE),Accuracy(ACC),Curve(AUC),and F1-score on three publicly available fundus datasets such as DRIVE,STARE,and CHASE_DB1.

关键词： Retinal vessel segmentation deformable convolution multi-scale transformer hybrid loss function

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

RealFuVSR:Feature enhanced real-world video super-resolution

引用

Virtual Reality & Intelligent Hardware 2023年第6期5卷 523-537页

作者： Zhi LI Xiongwen PANG Yiyue JIANG Yujie WANG School of Artificial Intelligence South China Normal UniversityGuangzhou 510631China School of Computer Science South China Normal UniversityGuangzhou 510631China School of Economics and Management South China Normal UniversityGuangzhou 510631China

Background Recurrent recovery is a common method for video super-resolution(VSR)that models the correlation between frames via hidden ***,the application of this structure in real-world scenarios can lead to unsatisfactory *** found that in real-world VSR training,the use of unknown and complex degradation can better simulate the degradation process in the real *** Based on this,we propose the RealFuVSR model,which simulates real-world degradation and mitigates artifacts caused by the ***,we propose a multiscale feature extraction module(MSF)module that extracts and fuses features from multiple scales,thereby facilitating the elimination of hidden state *** improve the accuracy of the hidden state alignment information,RealFuVSR uses an advanced optical flow-guided deformable ***,a cascaded residual upsampling module was used to eliminate noise caused by the upsampling *** The experiment demonstrates that RealFuVSR model can not only recover high-quality videos but also outperforms the state-of-the-art RealBasicVSR and RealESRGAN models.

关键词： Video super-resolution deformable convolution Cascade residual upsampling Second-order degradation Multi-scale feature extraction

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

HSCA-Net: A Hybrid Spatial-Channel Attention Network in Multiscale Feature Pyramid for Document Layout Analysis

引用

Journal of Artificial Intelligence and Technology 2023年第1期3卷 10-17页

作者： Honghong Zhang Canhui Xu Cao Shi Henyue Bi Yuteng Li Sami Mian Qingdao University of Science and Technology Qingdao 266000China University of Pittsburgh PittsburghUSA

Document images often contain various page components and complex logical structures,which make document layout analysis task *** most deep learning-based document layout analysis methods,convolutional neural networks(CNNs)are adopted as the feature extraction *** this paper,a hybrid spatial-channel attention network(HSCA-Net)is proposed to improve feature extraction capability by introducing attention mechanism to explore more salient properties within document *** HSCA-Net consists of spatial attention module(SAM),channel attention module(CAM),and designed lateral attention *** adaptively adjusts channel feature responses by emphasizing selective information,which depends on the contribution of the features of each *** guides CNNs to focus on the informative contents and capture global context information among page *** lateral attention connection incorporates SAM and CAM into multiscale feature pyramid network,and thus retains original feature *** effectiveness and adaptability of HSCA-Net are evaluated through multiple experiments on publicly available datasets such as PubLayNet,ICDAR-POD,and Article *** results demonstrate that HSCA-Net achieves state-of-the-art performance on document layout analysis task.

关键词： layout analysis attention mechanism deep learning deformable convolution

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Method to Appraise Dangerous Class of Building Masonry Component Based on DC-YOLO Model

引用

Computers, Materials & Continua 2020年第4期63卷 457-468页

作者： Hongrui Zhang Wenxue Wei Xinguang Xiao Song Yang Wanlu Shao College of Computer Science and Engineering Shandong University of Science and TechnologyQingdao266590China

This DC-YOLO Model was designed in order to improve the efficiency for appraising dangerous class of buildings and avoid manual intervention,thereby making the appraisal results more *** is an automated method designed based on deep learning and target detection algorithms to appraise the dangerous class of building masonry ***,it(1)adopted K-means clustering to obtain the quantity and size of the prior boxes;(2)expanded the grid size to improve identification to small targets;(3)introduced in deformable convolution to adapt to the irregular shape of the masonry component *** experimental results show that,comparing with the conventional method,the DC-YOLO model has better recognition rates for various targets to different extents,and achieves good effects in precision,recall rate and F1 value,which indicates the good performance in classifying dangerous classes of building masonry component.

关键词： Deep learning masonry component appraisal of dangerous class deformable convolution

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Research on Facial Fatigue Detection of Drivers with Multi-feature Fusion

引用

Instrumentation 2023年第1期10卷 23-31页

作者： YE Yuxuan ZHOU Xianchun WANG Wenyan YANG Chuanbin ZOU Qingyu School of Electronic and Information Engineering Nanjing University of Information Science and TechnologyNanjing 210044 School of Artificial Intelligence(School of Future Technology) Nanjing University of Information Science and TechnologyNanjing 210044

In order to solve the shortcomings of current fatigue detection methods such as low accuracy or poor real-time performance,a fatigue detection method based on multi-feature fusion is ***,the HOG face detection algorithm and KCF target tracking algorithm are integrated and deformable convolutional neural network is introduced to identify the state of extracted eyes and mouth,fast track the detected faces and extract continuous and stable target faces for more efficient *** the head pose algorithm is introduced to detect the driver’s head in real time and obtain the driver’s head state ***,a multi-feature fusion fatigue detection method is proposed based on the state of the eyes,mouth and *** to the experimental results,the proposed method can detect the driver’s fatigue state in real time with high accuracy and good robustness compared with the current fatigue detection algorithms.

关键词： HOG Face Posture Detection deformable convolution Multi-feature Fusion Fatigue Detection

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

欢迎您,

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

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

请选择保存的检索档案：

请选择收藏分类：

通借通还

欢迎您,

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

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

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：