检索结果-内蒙古大学图书馆

arXiv 2020年

作者： Yin, Junbo Shen, Jianbing Guan, Chenye Zhou, Dingfu Yang, Ruigang Beijing Lab of Intelligent Information Technology School of Computer Science Beijing Institute of Technology China Baidu Research National Engineering Laboratory of Deep Learning Technology and Application China Inception Institute of Artificial Intelligence United Arab Emirates University of Kentucky Kentucky United States

Existing LiDAR-based 3D object detectors usually focus on the single-frame detection, while ignoring the spatiotemporal information in consecutive point cloud frames. In this paper, we propose an end-to-end online 3D video object detector that operates on point cloud sequences. The proposed model comprises a spatial feature encoding component and a spatiotemporal feature aggregation component. In the former component, a novel Pillar Message Passing Network (PMPNet) is proposed to encode each discrete point cloud frame. It adaptively collects information for a pillar node from its neighbors by iterative message passing, which effectively enlarges the receptive field of the pillar feature. In the latter component, we propose an Attentive Spatiotemporal Transformer GRU (AST-GRU) to aggregate the spatiotemporal information, which enhances the conventional ConvGRU with an attentive memory gating mechanism. AST-GRU contains a Spatial Transformer Attention (STA) module and a Temporal Transformer Attention (TTA) module, which can emphasize the foreground objects and align the dynamic objects, respectively. Experimental results demonstrate that the proposed 3D video object detector achieves state-of-the-art performance on the large-scale nuScenes benchmark. Copyright © 2020, The Authors. All rights reserved.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

DCache: A Distributed Cache Mechanism for HDFS based on RDMA

DCache: A Distributed Cache Mechanism for HDFS based on RDMA

引用

IEEE International Conference on High Performance Computing and Communications (HPCC)

作者： Xiao Zhang Binbin Liu Zian Gou Jia Shi Xiaonan Zhao School of Computer Science Northwestern Polytechnical University 127 Youyi xi Road Xi’an Shaanxi China Key Laboratory of Big Data Storage and Management of MIIT in NWPU National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology in NWPU

ISBN: (纸本)9781728176505

Hadoop Distributed File System (HDFS) is an important component of Hadoop, which provides data storage service. The performance of the IO subsystem has a great influence on data processing efficiency. In the Hadoop system, computing jobs are scheduled to the nodes where the data is located to reduce IO time. The job scheduler and data distribution have a great impact on the efficiency of data processing. HDFS provides a mechanism for users to specify files or directories to be cached. But the cached data is only accessible for the jobs running in the same data nodes. In this paper, we present a distributed cache mechanism for HDFS with RDMA-capable network. We separate the caching function from the original data nodes. We design and implement a new kind of component for the caching function. The cached data in these nodes can be accessed directly by any nodes in the cluster through Remote Direct Memory Access (RDMA). Compared with the existing caching mechanism, our approach improves the performance and stability of the IO process. We also accelerate the data writing process using our caching mechanism. The experiments show that compared with the existing caching mechanism of HDFS, our scheme improves the reading latency up to 3 times, improves the throughput by 30%~95% for large files, and the writing performance by 64%. When the task runs on different nodes, the IO fluctuation caused by different data distribution is dropped from 20.71 to 0.50.

关键词： Fluctuations File systems High performance computing Conferences Memory Writing Throughput

来源：评论

学校读者我要写书评

暂无评论

Nanoarchitectonics and radiation mitigation in Sm 2 O 3 -doped lithium borotellurite glass systems: From amorphous harmony to localized order

引用

Journal of Non-Crystalline Solids 2025年 666卷

作者： G. Kilic Hasan Durmus Selçuk Birdogan E. Ilik Esra Kavaz Perişanoğlu Sevda Saltık Ömer Guler Ghada ALMisned H.O. Tekin Eskisehir Osmangazi University Faculty of Science Department of Physics TR-26040 Eskisehir Türkiye Vocational School of Health Services Izmir University of Economics Izmir Türkiye Sabanci University Nanotechnology Research and Application Center (SUNUM) 34956 Istanbul Türkiye Ataturk University Faculty of Science Department of Physics Erzurum Türkiye Munzur University Rare Earth Elements Application and Research Center 62000 Tunceli Türkiye Department of Physics College of Science Princess Nourah Bint Abdulrahman University P.O. Box 84428 Riyadh 11671 Saudi Arabia Medical Diagnostic Imaging Department College of Health Sciences University of Sharjah 27272 Sharjah United Arab Emirates Istinye University Computer Engineering Department Faculty of Engineering and Natural Sciences 34396 Istanbul Türkiye

In this study, a novel series of nano-Sm 2 O 3 -doped lithium borotellurite glasses was synthesized and systematically investigated in terms of their structural, physical, and radiation shielding characteristics. X-ray diffraction confirmed the amorphous nature of all samples, while transmission electron microscopy revealed progressive nanoarchitectural refinement and particle growth with increasing Sm 2 O 3 content, from uniformly dispersed nanodomains to dense agglomerates exhibiting short-range crystalline ordering. Energy-dispersive X-ray spectroscopy and elemental mapping validated the homogeneous incorporation of boron, oxygen, tellurium, and samarium throughout the glass matrix without phase separation. Gamma-ray shielding evaluations showed enhanced mass attenuation coefficients and reduced half-value layers, particularly in the Sm(n)8 sample, with significant performance gains at lower photon energies. Neutron shielding assessments further indicated elevated macroscopic removal cross-section values and dose attenuation, affirming the compositional benefit of Sm 2 O 3 for fast neutron mitigation. It can be concluded that nano-Sm 2 O 3 incorporation would play a multifunctional role on structural and radioprotective properties in addition to morphostructural features of lithium borotellurite glasses.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Fine-Grained Image Classification Combined with Label Description

Fine-Grained Image Classification Combined with Label Descri...

引用

International Conference on Tools for Artificial Intelligence (ICTAI)

作者： Xiruo Shi Liutong Xu Pengfei Wang School of Computer Science Beijing University of Posts and Telecommunications National Engineering Laboratory of Big Data Application on Improving Government Governance Capabilities Beijing China School of Computer Science Beijing University of Posts and Telecommunications Beijing China

Fine-grained image classification faces huge challenges because fine-grained images are similar overall, and the distinguishable regions are difficult to find. Generally, in this task, label descriptions contain valuable semantic information that is accurately compatible with discriminative features of images (i.e., the description of the "Rusty Black Bird" corresponding to the morphological characteristics of its image). Bringing these descriptions into consideration is benefit to discern these similar images. Previous works, however, usually ignore label descriptions and just mine informative features from images, thus the performance may be limited. In this paper, we try to take both label descriptions and images into consideration, and we formalize the classification task into a matching task to address this issue. Specifically, Our model is based on a combination of Convolutional Neural Networks (CNN) over images and Graph Convolutional Networks(GCN) over label descriptions. We map the resulting image representations and text representations to the same dimension for matching and achieve the purpose of classification through the matching operation. Experimental results demonstrate that our approach can achieve the best performance compared with the state-of-the-art methods on the datasets of Stanford dogs and CUB-200-2011.

关键词：

来源：评论

学校读者我要写书评

暂无评论

An approach to the applicability evaluation of moving target tracking algorithm 2nd

An approach to the applicability evaluation of moving target...

引用

2nd Chinese Conference on Pattern Recognition and computer Vision, PRCV 2019

作者： Xi, Runping Xue, Shaohui Han, Qianqian Chen, Jiaxin School of Computer Science Northwestern Polytechnical University Xi’an710129 China National Engineering Laboratory of Integrated Aero-Space-Ground-Ocean Big Data Application Technology Xi’an710129 China Shaanxi Provincial Key Laboratory of Speech and Image Information Processing Xi’an710129 China

ISBN: (纸本)9783030317256

Objective and effective algorithm performance evaluation results are an important basis for the selection of tracking algorithms. Problems in the existing performance evaluation of moving target tracking algorithms include an enlarge number of trials, and in particular, failure to consider the influence of algorithm performance on the multifactor combination scenario. This study proposes a method based on the orthogonal test to evaluate algorithms. First, the factors and levels of the tracking algorithm are analyzed, and an orthogonal test dataset is constructed by using an orthogonal table. Second, the experiments of the performance evaluation are arranged with the dataset and the results are analyzed via range analysis. Finally, evaluation results show that the strong– weak sequence of factors affect the performance of the algorithm and the combination of levels form the factors that can achieve enhanced algorithm performance. Experimental results show that the proposed method can evaluate algorithms comprehensively, objectively, and effectively with decreased test and data volume. © Springer Nature Switzerland AG 2019.

关键词： Clutter (information theory)

来源：评论

学校读者我要写书评

暂无评论

SPRSound: Open-Source SJTU Paediatric Respiratory Sound Database

TechRxiv

引用

TechRxiv 2022年

作者： Li, Yongfu Zhang, Qing Zhang, Jing Yuan, Jiajun Huang, Huajie Zhang, Yuhang Zhang, Baoqin Lv, Gaomei Lin, Shuzhu Wang, Na Liu, Xin Tang, Mingyu Wang, Yahua Ma, Hui Liu, Lu Yuan, Shuhua Zhou, Hongyuan Zhao, Jian Yin, Yong Zhao, Liebin Wang, Guoxing Lian, Yong Department of Micro-Nano Electronics and MoE Key Lab of Artificial Intelligence Shanghai Jiao Tong University Shanghai200240 China Pediatric AI Clinical Application and Research Center Shanghai Children’s Medical Center Shanghai Jiao Tong University School of Medicine China Department of Respiratory Medicine Shanghai Children’s Medical Center Shanghai Jiao Tong University School of Medicine China China School of Computer Engineering and Science Shanghai University China Sanya Maternity and Child Care Hospital China Department of Paediatrics Taicang Affiliated Hospital of Soochow University The First People’s Hospital of Taicang China Linyi City People Hospital China Department of Paediatrics Fengcheng Hospital China Department of Pediatrics Fifth Affiliated Hospital of Harbin Medical University China Shanghai Tuoxiao Intelligent Technology Co. Ltd China

It has proved that the auscultation of respiratory sound has advantage in early respiratory diagnosis. Various methods have been raised to perform automatic respiratory sound analysis to reduce subjective diagnosis and physicians’ workload. However, these methods highly rely on the quality of respiratory sound database. In this work, we have developed the first open-access paediatric respiratory sound database, SPRSound. The database consists of 2,683 records and 9,089 respiratory sound events from 292 participants. Accurate label is important to achieve a good prediction for adventitious respiratory sound classification problem. A custom-made sound label annotation software (SoundAnn) has been developed to perform sound editing, sound annotation, and quality assurance evaluation. A team of 11 experienced paediatric physicians is involved in the entire process to establish golden standard reference for the dataset. To verify the robustness and accuracy of the classification model, we have investigated the effects of different feature extraction methods and machine learning classifiers on the classification performance of our dataset. As such, we have achieved a score of 75.22%, 61.57%, 56.71%, and © 2022, CC BY.

关键词： Pediatrics

来源：评论

学校读者我要写书评

暂无评论

Constructing a Comprehensive Clinical Database Integrating Patients' Data from Intensive Care Units and General Wards 12

Constructing a Comprehensive Clinical Database Integrating P...

引用

12th International Congress on Image and Signal Processing, BioMedical engineering and Informatics, CISP-BMEI 2019

作者： Liu, Tongbo Liu, Xiaoli Fan, Yong Xu, Haoran Ng, Yeuk Lam Li, Peiyao Xue, Wanguo Zhang, Zhengbo Chinese PLA General Hospital Department of Computer Management and Application Beijing China School of Biological Science and Medical Engineering Beihang University Beijing China Chinese PLA General Hospital Department of Biomedical Engineering Beijing China University of Toronto Faculty of Arts Science Toronto Canada Medical Big Data Center Chinese PLA General Hospital Beijing China

ISBN: (纸本)9781728148526

Collection and analysis of large volumes of ICU data are invaluable to the advancement of clinical knowledge, and large-scale ICU databases have been effective resources to understand risk factors and perform predictive analysis by using machine learning. This paper introduces the construction of a comprehensive clinical database: PLAGH-ICU database, which integrated data from nine intensive care units, emergency department and general wards. Data from several sources were extracted and integrated in the database, including patient demographics, hospital administrative data, physiological data, medications, lab test, fluid balance data, notes and reports etc. Detailed information about the database, such as patient characteristics, disease distribution, category of data, data records were illustrated. As far as we know, this is the first comprehensive ICU database developed and used as a research database in datathon event in China. Such kind of database will promote the research progress in critical care medicine, as well as the development of modern hospital information system in China. © 2019 IEEE.

关键词： Intensive care units

来源：评论

学校读者我要写书评

暂无评论

Online Optimal Event-triggered Tracking Control with Actuator Saturation via ADP

Online Optimal Event-triggered Tracking Control with Actuato...

引用

Data Driven Control and Learning Systems (DDCLS)

作者： Lu Liu Ruizhuo Song Yi Ren Yong Zhao Biao Song School of Automation and Electrical Engineering University of Science and Technology Beijing Beijing P. R. China China communications construction intelligence transportation company limited Tianjin P. R. China Beijing Institute of Computer Application University of Science and Technology Beijing Beijing P. R. China

ISBN: (数字)9781728159225

ISBN: (纸本)9781728159232

This paper comes up with an optimal tracking control mechanism with saturation controller for some continuous nonlinear systems based on the event-triggered (ET) mechanism via adaptive dynamic programming (ADP). To begin with, the constrained tracking control problems are able to be transformed into common event-triggered optimal control through designing an appropriate cost function. Next, a single critic network is to estimate the cost function. Moreover, taking the appropriate triggering condition as the prerequisite, the stability proof of the system is proved according to Lyapunov theory. In addition, the simulations verify the excellent performance of ET-ADP algorithm because the optimal control law is only updated at the required times influenced by the devised triggering condition, which can accomplish high efficiency of resources and energy consumption of the system.

关键词： Optimal control Dynamic programming Cost function Nonlinear systems Mathematical model Heuristic algorithms Feeds

来源：评论

学校读者我要写书评

暂无评论

application of NARX dynamic neural network in blood glucose prediction model

Application of NARX dynamic neural network in blood glucose ...

引用

Data Driven Control and Learning Systems (DDCLS)

作者： Menglin Xu Ruizhuo Song Yong Zhao Biao Song Juan Tang School of Automation and Electrical Engineering University of Science and Technology Beijing Beijing P. R. China Beijing Institute of Computer Application University of Science and Technology Beijing Beijing P. R. China Ericsson (China) Commodity Co. Ltd ARIIX CHINA P. R. China

ISBN: (数字)9781728159225

ISBN: (纸本)9781728159232

Establishing a blood glucose prediction model and controlling the amount of insulin injected according to the predicted blood glucose value can effectively reduce the harm caused by high and low blood glucose to the human body. Because the dynamic process of glucose-insulin metabolism in human body is a typical non-steady-state and non-linear process, and has a large time-lag dynamic characteristic, it is difficult for linear models to accurately describe the blood glucose-insulin model. With the development of artificial neural network and its application in all aspects of modern times, this paper proposes to use NARX dynamic neural network to build blood glucose-insulin model, and use the data generated by UVa/Padova simulation platform for network training. Finally, through MATLAB simulation verification, by calculating the mean square error and Clark error grid analysis method, it is shown that the blood glucose can be effectively predicted.

关键词： Blood Glucose Neural networks Data models Predictive models Insulin Mathematical model

来源：评论

学校读者我要写书评

暂无评论

Gaussian noise removal with exponential functions and spectral norm of weighted hankel matrices

arXiv

引用

arXiv 2020年

作者： Qiu, Tianyu Liao, Wenjing Guo, Di Liu, Dongbao Wang, Xin Qu, Xiaobo Department of Electronic Science Fujian Provincial Key Laboratory of Plasma and Magnetic Resonance Xiamen University Xiamen China School of Mathematics Georgia Institute of Technology AtlantaGA United States School of Computer and Information Engineering Fujian Provincial University Key Laboratory of Internet of Things Application Technology Xiamen University of Technology Xiamen China

Exponential functions are powerful tools to model signals in various scenarios, such as magnetic resonance spectroscopy/ imaging, radar, and concatenative text-to-speech synthesis. Exponential signals, however, are usually corrupted by Gaussian noise in practice, raising difficulties in sequential analysis and quantification of the signals. In this work, we propose a denoising method based on low-rank Hankel matrices for exponential signals corrupted by Gaussian noise. An accurate estimate on the spectral norm of weighted Hankel matrices is provided as a theoretical guidance to set the regularization parameter. The bound can be efficiently calculated since it only depends on the standard deviation of the noise and a constant. Aided by the bound, one can easily obtain a good regularization parameter to produce promising denoised results. Our experiments on simulated and magnetic resonance spectroscopy data demonstrate a superior denoising performance of our proposed approach in comparison with the typical Cadzow and the state-ofthe-art QR decomposition methods, especially in the low signalto-noise ratio regime. Copyright © 2020, The Authors. All rights reserved.

关键词： Matrix algebra

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：