检索结果-内蒙古大学图书馆

13th International Conference on Computing and Pattern Recognition, ICCPR 2024

作者： Li, Zongshu Zhu, Guibo Yi, Dongyi Wang, Jinqiao Foundation Model Research Center Institute of Automation Chinese Academy of Sciences Beijing China Algorithm Development Department Wuhan Artificial Intelligence Research Wuhan Hubei China

ISBN: (纸本)9798400717482

Mamba - the selective SSM (S6) model, makes up for inherent shortcomings of Transformer in long sequence processing and has been gaining significant attention from researchers. In this paper, we employ Mamba to build a multimodal chat model - Multimodal Mamba. Inspired by the prevalent structure of multimodal models, MM-Mamba consists of three parts: vision encoder, multimodal projector and language decoder. Firstly, we simply replace the text decoder from Transformer to Mamba. Secondly, considering that parameters and pretraining tokens of Mamba are significantly less than Transformer, we propose a multi-stage training pipeline to train MM-Mamba. Finally, in order to further boost the proficiency of MM-Mamba across a wide range of multimodal downstream tasks, we introduce several powerful techniques, such as higher input image resolution, noisy embedding finetuning. Compared with multimodal large language models (MLLMs) built upon Transformer, MM-Mamba can be adapted to various downstream tasks with extremely low temporal and economic training cost. Experimental results showed that albeit with notably fewer parameters and significantly limited training data volume, remarkable results are achieved by MM-Mamba not only on English benchmarks but also in Chinese downstream tasks. Code is released at here. © 2024 Copyright held by the owner/author(s).

关键词： Mamba

来源：评论

学校读者我要写书评

暂无评论

Decomposition and comparative analysis of urban-rural differences in eHealth literacy among nursing students: A multisite cross-sectional study

引用

Nurse Education Today 2025年 146卷 106562页

作者： Wang, Xue Ge, Lina Yu, Zhenjie Qiao, Qiao Guo, Xiaoming Liu, Shiping Zhang, Chengcai Zang, Shuang Department of Community Nursing School of Nursing China Medical University Shenyang China Department of Obstetrics and Gynecology Shengjing Hospital of China Medical University Shenyang China School of Nursing Tianjin Medical University Tianjin China Department of Radiation Oncology The First Hospital of China Medical University Shenyang China Algorithm Research and Development Department Neusoft IntelliRay Technology Co. Ltd. Shenyang China

Background: Investigating the urban-rural differences in eHealth literacy among Chinese nursing students is imperative for enhancing healthcare education and practice. This study aimed to investigate the factors that contribute to urban-rural differences in eHealth literacy among Chinese nursing students and measure their respective contributions. Design: A multisite cross-sectional study. Settings: In China. Participants: A total of 967 nursing students were involved in this study. Methods: Multivariate linear regression models were used to explore the associations between sociodemographic and health-related factors and the eHealth literacy of Chinese nursing students. Additionally, Blinder-Oaxaca decomposition was employed to scrutinize the factors attributing to the urban-rural differences in eHealth literacy among Chinese nursing students, elucidating their respective contributions. Results: The mean eHealth literacy score for urban nursing students (mean: 20.26, SD: 3.61) exceeded that of rural nursing students (mean: 19.53, SD: 3.56) (p < 0.001). Blinder-Oaxaca decomposition analysis indicated that 84.82 % of the urban-rural differences in eHealth literacy among nursing students can be explained. The urban-rural differences are primarily explained by family health (40.05 %). Conclusions: The study emphasizes the importance of tailoring educational curricula to address urban-rural differences in eHealth literacy among nursing students in China. Specifically, incorporating content related to family health in educational programs is crucial. © 2024 Elsevier Ltd

关键词： China Comparative analysis Decomposition analysis eHealth literacy Nursing students Urban-rural difference

来源：评论

学校读者我要写书评

暂无评论

DM Code Key Point Detection algorithm Based on CenterNet

引用

Computers, Materials & Continua 2023年第11期77卷 1911-1928页

作者： Wei Wang Xinyao Tang Kai Zhou Chunhui Zhao Changfa Liu School of Information Engineering Chang’an UniversityXi’an710018China Algorithm Research and Development Department GRGBanking Equipment Co.Ltd.Guangzhou510663China Research and Development Department Xi’an Soar Electromechanical TechnologyLtd.Xi’an710043China

Data Matrix(DM)codes have been widely used in industrial *** reading of DM code usually includes positioning and *** positioning is a prerequisite for successful *** image processing methods have poor adaptability to pollution and complex *** deep learning-based methods can automatically extract features,the bounding boxes cannot entirely fit the contour of the *** image processing methods are required for precise positioning,which will reduce *** of the above problems,a CenterNet-based DM code key point detection network is proposed,which can directly obtain the four key points of the DM *** with the existing methods,the degree of fitness is higher,which is conducive to direct *** further improve the positioning accuracy,an enhanced loss function is designed,including DM code key point heatmap loss,standard DM code projection loss,and polygon Intersection-over-Union(IoU)loss,which is beneficial for the network to learn the spatial geometric characteristics of DM *** experiment is carried out on the self-made DM code key point detection dataset,including pollution,complex background,small objects,etc.,which uses the Average Precision(AP)of the common object detection metric as the evaluation *** reaches 95.80%,and Frames Per Second(FPS)gets 88.12 on the test set of the proposed dataset,which can achieve real-time performance in practical applications.

关键词： DM code key point detection CenterNet object detection enhanced loss function

来源：评论

学校读者我要写书评

暂无评论

Deterministic and Stochastic Simulators for Non-Isotropic V2V-MIMO Wideband Channels

引用

China Communications 2018年第7期15卷 18-29页

作者： Yiran Li Xiang Cheng Nan Zhang State Key Laboratory of Advanced Optical Communication Systems and Networks School of Electronics Engineering and Computer Science Peking University Wireless Algorithm Department Product Research and Development System ZTE Corporation

In this paper, we consider a novel two-dimensional(2D) geometry-based stochastic model(GBSM) for multiple-input multiple-output(MIMO) vehicle-to-vehicle(V2V) wideband fading channels. The proposed model employs the combination of a two-ring model and a multiple confocal ellipses model, where the signal is sum of the line-of-sight(Lo S) component, single-bounced(SB) rays, and double-bounced(DB) rays. Based on the reference model, we derive some expressions of channel statistical properties, including space-time correlation function(STCF), Doppler spectral power density(DPSD), envelope level crossing rate(LCR) and average fade duration(AFD). In addition, corresponding deterministic and stochastic simulation models are developed based on the reference model. Moreover, we compare the statistical properties of the reference model and the two simulation models in different scenarios and investigate the impact of different vehicular traffic densities(VTDs) on the channel statistical properties of the proposed model. Finally, the great agreement between simulation models and the reference model demonstrates not only the utility of simulation models, but also the correctness of theoretical derivations and simulations.

关键词： vehicle-to-vehicle wideband channels simulation model statistical properties

来源：评论

学校读者我要写书评

暂无评论

Field Trial of End-to-end Management and Control of Semiactive WDM System for 5G Centralized Front-haul Network

Field Trial of End-to-end Management and Control of Semiacti...

引用

2023 Optical Fiber Communications Conference and Exhibition, OFC 2023

作者： Zhao, Yang Sun, Jiang Zheng, Haomian Yu, Chaode Wang, Dong Cai, Qian Li, Yunbo Han, Liuyan Zhang, Dechao Li, Han China Mobile Research Institute Department of Network and It Technology Beijing100053 China Huawei Technologies Co. Ltd Optical System & Algorithm Development Department Wuhan430223 China

A novel management and control architecture and abstract model of semi-active WDM system for 5G front-haul network are proposed. The functions of device information acquisition, alarm monitoring and fault location is ... 详细信息

ISBN: (纸本)9781957171180

关键词： 5G mobile communication systems

来源：评论

学校读者我要写书评

暂无评论

Field Trial of a New Metropolitan WDM Architecture and Flexible Scheduling based on Optical Layer OAM

Field Trial of a New Metropolitan WDM Architecture and Flexi...

引用

2024 Asia Communications and Photonics Conference, ACP 2024 and International Conference on Information Photonics and Optical Communications, IPOC 2024

作者： Zhao, Yang Liu, Yucong Li, Xiang Wang, Dong Hu, Changbin Zhang, Dechao Department of Fundamental Network Technology China Mobile Research Institute Beijing China Optical System & Algorithm Development Department Huawei Technologies Co. Ltd. Wuhan China

A novel wavelength pooling technology and a flexible scheduling method based on optical layer OAM (O-OAM) are proposed. Costs of network construction and maintenance are reduced significantly, and service latency is r... 详细信息

ISBN: (纸本)9798350379266

关键词： Centralized Control Metropolitan Transport Network Optical Layer OAM Wavelength Pooling

来源：评论

学校读者我要写书评

暂无评论

A Large Scale Extended algorithm for 2D Halton Points with Low-Discrepancy Sequences 3

A Large Scale Extended Algorithm for 2D Halton Points with L...

引用

3rd International Conference on Artificial Intelligence and Advanced Manufacture, AIAM 2021

作者： Chen, Wenxing Dai, Shuyang Zheng, Baojuan Wuhan University School of Mathematics and Statistics Wuhan430072 China Commercial Vehicle Technology Center of SAIC Motor Group Co. Ltd Department of Algorithm Research and Development Shanghai200433 China

ISBN: (纸本)9781665417327

Random discrete points have important application value in meshless PDE equation discretization, molecular dynamics simulation, point cloud imaging and so on. There are many common methods to generate random points, such as Monte Carlo, Gibbs Sampling, and Hammersley series and etc. But Halton random point algorithm has a defect that it only generates discrete points in [0, 1]2 region. However, in practical applications, it is necessary to be able to generate discrete points on any area. This paper proposed a new Halton points extension algorithm to solve this defects. We defined a linear operator which can transform discrete points from [0, 1]2 region into any plane region. Two examples are given, the extension algorithm respectively includes square, rectangular and polar coordinates region. The numerical results show that our method is accurate, effective and more general, it also enhanced the application range by our method. © 2021 IEEE.

关键词： Molecular dynamics

来源：评论

学校读者我要写书评

暂无评论

The Control Principle and Mathematical Model of Intelligent Solar Panels

The Control Principle and Mathematical Model of Intelligent ...

引用

2021 International Conference on Robotics Automation and Intelligent Control, ICRAIC 2021

作者： Chen, Wenxing Dai, Shuyang Zheng, Baojuan Tang, Zhuyan Lin, Hao School of Mathematics and Statistics Wuhan University Wuhan430072 China Department of Algorithm Research and Development Commerical Vehicle Technology Center SAIC Motor Group Co.Ltd Shanghai200433 China

Energy is an important material that promotes social development and maintains human life. Clean energy has become a scientific research theme of open world which appeal to many scientists to research it. This paper inspired by the experience of sunflower growth. So we proposed a mathematical model for automatic adjustment of the maximum luminous flux for a limited solar panel, which can help to adjust the best light position. The physical model added to the light detector in the practical application, real-time comparison of the difference in illumination at different positions to adjust the direction of solar panel, thereby enhancing the utilization of light energy. In addition, in order to save space and prevent severe weather attacks, we designed a autofolding circle disc and retractable rectangular solar panels. The intelligent system judgement and decision making based on the electrical signals fed back by the wind sensor or the vibration sensor. When the wind speed reaches the threshold, it can be folded or contracted automatically, reducing manual maintenance. The fully automatic solar panels are designed in this paper, the research results show that theis model has a high utilization, strong sensitivity, safety and durability, and easy control. This is also a process of theoretical exploration of interdisciplinary services for the energy system. © 2022 Institute of Physics Publishing. All rights reserved.

关键词： Solar energy

来源：评论

学校读者我要写书评

暂无评论

Deep transfer learning via minimum enclosing balls 25th

Deep transfer learning via minimum enclosing balls

引用

25th International Conference on Neural Information Processing, ICONIP 2018

作者： Deng, Zhilong Liu, Fan Zhao, Jiangjiang Wei, Qiang Pang, Shaoning Leng, Yue Algorithm Group Research and Development Department Online Services of China Mobile Beijing China Artificial Intelligence Lab Meituan-Dianping Group Beijing China Department of Computing Unitec Institute of Technology Private Bag 92025 Auckland1142 New Zealand Algorithm Group Artificial Intelligence Lab Emotibot Technologies Limited Beijing China

ISBN: (纸本)9783030041816

Training of deep learning algorithms such as CNN, LSTM, or GRU often requires large amount of data. However in real world applications, the amount of data especially labelled data is limited. To address this challenge, we study Deep Transfer Learning (DTL) in the context of Multitasking Learning (MTL) to extract sharable knowledge from tasks and use it for related tasks. In this paper, we use Minimum Closed Ball (MEB) as a flexible knowledge representation method to map shared domain knowledge from primary task to secondary task in multitasking learning. The experiments provide both analytic and empirical results to show the effectiveness and robustness of the proposed MEB-based deep transfer learning. © Springer Nature Switzerland AG 2018.

关键词： Learning algorithms

来源：评论

学校读者我要写书评

暂无评论

An Efficient Evaluation Method for Automobile Shells Design Based on Semi-supervised Machine Learning Strategy 2

An Efficient Evaluation Method for Automobile Shells Design ...

引用

2nd International Conference on Computer, Big Data and Artificial Intelligence, ICCBDAI 2021

作者： Chen, Wenxing Dai, Shuyang Zheng, Baojuan Lin, Hao School of Mathematics and Statistics Wuhan University Wuhan430072 China Department of Algorithm Research and Development Commerical Vehicle Technology Center of SAIC Motor Group Co. Ltd Shanghai 200433 China

Automobile is one of the important modes of transportation for human travel in today's society. Batch production in various countries in the world has also promoted the transformation of production concepts. At present, the development of the automobile industry is developing towards the trend of intelligence, personalizat-ion and sharing. Car appearance in a variety of ways, not every design is reasonable. Therefore, the main purpose of this article is to establish a scientific evaluation standard in order to large-scale test the quality of a variety of car shells design. The scientific nature is mainly reflected in combination the fluid-solid coupling knowledge and machine learning in this article, which can analyze the force of different shells in the flow field, and put out the cloud map information such as the stress, pressure and velocity of the shell. At last, analyze the best test samples and store them in the database, and then using semi-supervised heuristic algorithm to perform the sample training, the ultimate goal is to make the evaluation system more robust. The trained model can correctly evaluate each personalized car shape and give a reasonable score, which is convenient for car manufacturers to make best decision with personalized demand and scientific production. © 2022 Institute of Physics Publishing. All rights reserved.

关键词： Automotive industry

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：