检索结果-内蒙古大学图书馆

3-D contour-aware U-Net for efficient rectal tumor segmentation in magnetic resonance imaging

Medical engineering and Physics 2025年 140卷 104352-104352页

作者： Lu, Yi Dang, Jun Chen, Junzhang Wang, Yuanyuan Zhang, Tao Bai, Xiangzhi Image Processing Center Beihang University Beijing102206 China National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital & Shenzhen Hospital Chinese Academy of Medical Sciences and Peking Union Medical College Shenzhen518116 China Department of Oncology The First Affiliated Hospital of Chongqing Medical University Chongqing40016 China The State Key Laboratory of Virtual Reality Technology and Systems Beihang University Beijing100191 China Beijing Advanced Innovation Center for Biomedical Engineering Beihang University Beijing100083 China Key Laboratory of Spacecraft Design Optimization and Dynamic Simulation Technology Ministry of Education Beihang University Beijing102206 China

Magnetic resonance imaging (MRI), as a non-invasive detection method, is crucial for the clinical diagnosis and treatment plan of rectal cancer. However, due to the low contrast of rectal tumor signal in MRI, segmentation is often inaccurate. In this paper, we propose a new three-dimensional rectal tumor segmentation method CAU-Net based on T2-weighted MRI images. The method adopts a convolutional neural network to extract multi-scale features from MRI images and uses a Contour-Aware decoder and attention fusion block (AFB) for contour enhancement. We also introduce adversarial constraint to improve augmentation performance. Furthermore, we construct a dataset of 108 MRI-T2 volumes for the segmentation of locally advanced rectal cancer. Finally, CAU-Net achieved a DSC of 0.7112 and an ASD of 2.4707, which outperforms other state-of-the-art methods. Various experiments on this dataset show that CAU-Net has high accuracy and efficiency in rectal tumor segmentation. In summary, proposed method has important clinical application value and can provide important support for medical image analysis and clinical treatment of rectal cancer. With further development and application, this method has the potential to improve the accuracy of rectal cancer diagnosis and treatment. © 2025 IPEM

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Distribution and rendering of real-time scene based on the virtual reality system

Communications in Computer and Information Science

引用

Communications in Computer and Information Science 2014年 437卷 192-202页

作者： Zhang, Chengfang Wang, Guoping Wang, Guoping Shenzhen Graduate School Peking University Shenzhen China Graphics and Interaction Lab of EECS Peking University Beijing China Beijiing Engineering Technology Research Center of Virtual Simulation and Visualization Beijing China

ISBN: (纸本)9783662454978

There are differences between real and virtual scene information displayed on a virtual reality platform, not a true reflection of the reality of the real-time information. On the basis of virtual reality platform, how to collect, organize and publish the corresponding real-time virtual reality scene scenes information is becoming a new problem. In this essay, we will explain how to input, distribute and display real scene information based on the virtual reality system, a distributed virtual reality system. The current real-time video information corresponded with the specified scene object in the virtual reality system is recorded by the client, released to the streamer server and then distributed via content distribution server. The live streaming information displayed on the virtual reality client increases the fidelity and real-time of the virtual reality. The system is timely published, scalable and is capable of supporting remote deployment and distribution. The goal is to provide a reliable and effective real scene information dissemination and presentation of the real scene platforms based on the virtual reality environment. © Springer-Verlag Berlin Heidelberg 2014.

关键词： virtual reality

来源：评论

学校读者我要写书评

暂无评论

An efficient spatial anti-aliasing method for deferred shading

引用

Computer Aided Drafting,Design and Manufacturing 2017年第2期27卷 57-63页

作者： Ye Chengchun Lai Shunnan Li Sheng Peking University Shenzhen Graduate School Beijing Engineering Technology Research Center of Virtual Simulation and Visualization (Peking University) School of Electronics Engineering and Computer Science Peking University Peking University Information Technology Institute (Tianjin Binhai)

In recent years, image-based anti-aliasing techniques have been widely investigated in real-time computer graphics. Compared to traditional routine of anti-aliasing, image-based methods, which combined with deferred shading, can efficiently decouple anti-aliasing step from standard graphics pipeline. However, most of the existing image-based methods consume video memory, bandwidth and computation resources heavily. In this paper, we propose an optimized anti-aliasing method, which can efficiently reduce the multi-sampling rate by a multi-sampling mask. We carefully designed the rendering system to avoid serial execution on branch divergence, by utilizing both modern GPU’s features and latest shader model. Experimental results demonstrate that our method can achieve high-quality synthesized image with real-time performance.

关键词： anti-aliasing deferred shading compute shader execution divergence

来源：评论

学校读者我要写书评

暂无评论

Classroom Attendance Auto-management Based on Deep Learning

Classroom Attendance Auto-management Based on Deep Learning

引用

2017 2nd International Conference on Education,Sports,Arts and Management engineering(ICESAME 2017)

作者： Dan Wang Rong Fu Zuying Luo College of Information Science and Technology Beijing Normal University Research Center of Virtual Reality and Visualization Technology and Engineering Beijing Normal University

Attendance is an important part of daily classroom evaluation. This paper develops an automatic attendance system by integrating two deep learning algorithm Faster R-CNN face detection algorithm and Seeta Face face recognition algorithm. The results of numerous experiments indict that:(1) the system can record such five violations of classroom, that is absence, later arrival, early departure, free access, and carelessness for attendance, and give the attendance table which can reflect the learning situation of all students after school.(2) For small classrooms with length of less than 6 meters, 1080P classroom monitoring video can meet the needs of classroom attendance;but for 9 meters of big classrooms, you should use 4 K classroom surveillance video.

关键词： classroom evaluations attendance face detection face recognition

来源：评论

学校读者我要写书评

暂无评论

High performance flow field visualization with high-order access dependencies

High performance flow field visualization with high-order ac...

引用

IEEE Scientific visualization Conference (SciVis)

作者： Jiang Zhang Hanqi Guo Xiaoru Yuan Key Laboratory of Machine Perception (Ministry of Education) Peking University Argonne National Laboratory Beijing Engineering Technology Research Center of Virtual Simulation and Visualization Peking University

We present a novel model based on high-order access dependencies for high performance pathline computation in flow field. The high-order access dependencies are defined as transition probabilities from one data block to other blocks based on a few historical data accesses. Compared with existing methods which employed first-order access dependencies, our approach takes the advantages of high order access dependencies with higher accuracy and reliability in data access prediction. In our work, high-order access dependencies are calculated by tracing densely-seeded pathlines. The efficiency of our proposed approach is demonstrated through a parallel particle tracing framework with high-order data prefetching. Results show that our method can achieve higher data locality than the first-order access dependencies based method, thereby reducing the I/O requests and improving the efficiency of pathline computation in various applications.

关键词： Prefetching Data visualization Computational modeling Data models Electronic mail Scalability Computational fluid dynamics

来源：评论

学校读者我要写书评

暂无评论

A bottom-up scheme for user-defined feature exploration in vector field ensembles

A bottom-up scheme for user-defined feature exploration in v...

引用

IEEE Scientific visualization Conference (SciVis)

作者： Richen Liu Hanqi Guo Xiaoru Yuan Key Laboratory of Machine Perception (Ministry of Education) Peking University Argonne National Laboratory Beijing Engineering Technology Research Center of Virtual Simulation and Visualization Peking University

Most of the existing approaches to visualize vector field ensembles are achieved by visualizing the uncertainty of individual variables from different simulation runs. However, the comparison of the derived feature or user-defined feature, such as the vortex in ensemble flow is also of vital significance since they often make more sense according to the domain knowledge. In this work, we present a framework to extract user-defined feature from different simulation runs. Specially, we use a bottom-up searching scheme to help to extract vortex with a user-defined shape, and further compute the geometry information including the size, and the geo-spatial location of the extracted vortex. Finally we design some linked views to compare the feature between different runs.

关键词： Three-dimensional displays Meteorology visualization Feature extraction Computational modeling Data models Data visualization

来源：评论

学校读者我要写书评

暂无评论

Automatic Evaluation of Teacher-Student Interaction Based on Dialogue Text

Automatic Evaluation of Teacher-Student Interaction Based on...

引用

2017 2nd International Conference on Education,Sports,Arts and Management engineering(ICESAME 2017)

作者： Jiao Lu Dan Wang Zuying Luo College of Information Science and Technology Beijing Normal University Research Center of Virtual Reality and Visualization Technology and Engineering Beijing Normal University

Quantitative evaluation in class can effectively improve the quality of classroom teaching, but the automatic evaluation tool is very scarce. In order to be able to carry out large-scale automatic evaluation of teacher-student interaction, this paper first designed a set of indicator system for the classroom interactive evaluation;and then input teacher and student dialogue text from the classroom teaching video;and asked two education evaluation professionals to evaluate the interaction between teachers and students on the same lesson according to the video and text respectively to observe the validity of the interactive evaluation of teachers and students based on the dialogue text. This paper then presented a teacher-student interactive automatic coding technique(SAC) based on dialogue sentences, and constructed a set of classroom-teaching automatic evaluation system(BIAS) for teacher-student interaction. The experimental results show that BIAS not only provides more interactive evaluation function of teachers and students, but also improves the educational evaluation personnel's coding efficiency by 7.3 times compared with the Flanders coding system(FIAS), which is manually coded.

关键词： classroom observation teacher-student interaction text analysis classroom-teaching automatic evaluation

来源：评论

学校读者我要写书评

暂无评论

Constructing Multiple Scene Graphs in Distributed Environment

引用

Computer Aided Drafting,Design and Manufacturing 2015年第1期25卷 16-21页

作者： XIA Rui WANG Guo-ping LI Sheng WANG Heng Shenzhen Key Lab for Cloud Computing Technology and Applications Shenzhen Graduate School of Peking University Graphics & Interaction Lab. School of Electronics Engineering and Computer SciencesPeking University Beijing Engineering Technology Research Center of Virtual Simulation and Visualization Peking University

Scene graph is a infrastructure of the virtual reality system to organize the virtual scene with abstraction, it can provide facility for the rendering engine and should be integrated effectively on demand into a real-time system, where a large quantities of scene objects and resources can be manipulated and managed with high flexibility and reliability. We present a new scheme of multiple scene graphs to accommodate the features of rendering engine and distributed systems. Based upon that, some other functions, e.g. block query, interactive editing, permission management, instance response, ＂redo＂ and ＂undo＂, are implemented to satisfy various requirements. At the same time, our design has compatibility to popular C/S architecture with good concurrent performance. Above all, it is convenient to be used for further development. The results of experiments including responding time demonstrate its good performance.

关键词： multiple scene graphs authority scene editing distributed systems

来源：评论

学校读者我要写书评

暂无评论

Hierarchical Mesh Segmentation Based on Quadric Surface Fitting

Hierarchical Mesh Segmentation Based on Quadric Surface Fitt...

引用

IEEE International Conference on Computer-Aided Design and Computer Graphics

作者： Huijuan Zhang Chong Li Leilei Gao Guoping Wang Graphics and Interaction Lab Peking University Beijing China Peking University Beijing Beijing CN Beijing Engineering Technology Research Center of Virtual Simulation and Visualization Peking University Beijing China Cooperative Medianet Innovation Center Beijing China

In this paper we introduce a hierarchical region decomposition algorithm for triangle meshes segmentation based on quadric surface fitting. Initially, the whole mesh represents a single cluster. At every iteration, the region with largest error is divided and Lloyd iteration clustering is executed following, which make sure the segmentation has the minimum error in each hierarchical level. An error control strategy is used to get finally result automatically. Besides, users also can interpose this process by specifying the segmentation level. Each segmentation region after this process has a best fitting proxy of quadric surface. We propose an improved L 2,1 distance into a hybrid error, which is used in least square fit of quadric function. By comparing with the state-of-the art methods, our method is testified to be insensitive to noise. Model optimization and hole filling based on our algorithm are demonstrated in the last.

关键词： Solid modeling Clustering algorithms Surface treatment Design automation Surface fitting Computational modeling Measurement

来源：评论

学校读者我要写书评

暂无评论

Comparative visualization of vector field ensembles based on longest common subsequence

Comparative visualization of vector field ensembles based on...

引用

Pacific (formerly Asia-Pacific APVIS) visualization Symposium

作者： Richen Liu Hanqi Guo Jiang Zhang Xiaoru Yuan Key Laboratory of Machine Perception (Ministry of Education) Peking University Mathematics and Computer Science Division Argonne National Laboratory Beijing Engineering Technology Research Center of Virtual Simulation and Visualization Peking University

We propose a longest common subsequence (LCSS)-based approach to compute the distance among vector field ensembles. By measuring how many common blocks the ensemble pathlines pass through, the LCSS distance defines the similarity among vector field ensembles by counting the number of shared domain data blocks. Compared with traditional methods (e.g., pointwise Euclidean distance or dynamic time warping distance), the proposed approach is robust to outliers, missing data, and the sampling rate of the pathline timesteps. Taking advantage of smaller and reusable intermediate output, visualization based on the proposed LCSS approach reveals temporal trends in the data at low storage cost and avoids tracing pathlines repeatedly. We evaluate our method on both synthetic data and simulation data, demonstrating the robustness of the proposed approach.

关键词： Data models Data visualization Time series analysis Uncertainty Computational modeling Shape Measurement

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：