检索结果-南通市图书馆

EmotionIC: emotional inertia and contagion-driven dependency modeling for emotion recognition in conversation

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第8期67卷 130-146页

作者： yingjian liu Jiang LI Xiaoping WANG Zhigang ZENG School of Artificial Intelligence and Automation Huazhong University of Science and Technology Hubei Key Laboratory of Brain-inspired Intelligent Systems Huazhong University of Science and Technology Key Laboratory of Image Processing and Intelligent Control (Huazhong University of Science and Technology) Ministry of Education Institute of Artificial Intelligence Huazhong University of Science and Technology

Emotion recognition in conversation(ERC) has attracted growing attention in recent years as a result of the advancement and implementation of human-computer interface technologies. In this paper,we propose an emotional inertia and contagion-driven dependency modeling approach(EmotionIC) for ERC tasks. Our EmotionIC consists of three main components, i.e., identity masked multi-head attention(IMMHA), dialogue-based gated recurrent unit(DiaGRU), and skip-chain conditional random field(SkipCRF).Compared to previous ERC models, EmotionIC can model a conversation more thoroughly at both the feature-extraction and classification levels. The proposed model attempts to integrate the advantages of attention-and recurrence-based methods at the feature-extraction level. Specifically, IMMHA is applied to capture identity-based global contextual dependencies, while Dia GRU is utilized to extract speaker-and temporal-aware local contextual information. At the classification level, SkipCRF can explicitly mine complex emotional flows from higher-order neighboring utterances in the conversation. Experimental results show that our method can significantly outperform the state-of-the-art models on four benchmark *** ablation studies confirm that our modules can effectively model emotional inertia and contagion.

关键词： emotion recognition in conversation emotional inertia and contagion multi-head attention gated recurrent unit conditional random field

Three-Dimensional Ocean Sensor Networks:A Survey

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Journal of Ocean University of China 2012年第4期11卷 436-450页

作者： WANG Yu liu yingjian GUO Zhongwen Department of Computer Science University of North Carolina at Charlotte Charlotte North Carolina 28223 USA Department of Computer Science and Technology Ocean University of China Qingdao 266100 P. R. China

The past decade has seen a growing interest in ocean sensor networks because of their wide applications in marine research,oceanography,ocean monitoring,offshore exploration,and defense or homeland *** sensor networks are generally formed with various ocean sensors,autonomous underwater vehicles,surface stations,and research *** make ocean sensor network applications viable,efficient communication among all devices and components is *** to the unique characteristics of underwater acoustic channels and the complex deployment environment in three dimensional(3D) ocean spaces,new efficient and reliable communication and networking protocols are needed in design of ocean sensor *** this paper,we aim to provide an overview of the most recent advances in network design principles for 3D ocean sensor networks,with focuses on deployment,localization,topology design,and position-based routing in 3D ocean spaces.

关键词： ocean sensor networks underwater sensor networks three-dimensional sensor networks ocean applications 3D de-ployment topology design localization position-based routing

Applications of optically and electrically driven nanoscale bowtie antennas

在线全文

学校读者我要写书评

暂无评论

Opto-Electronic Science 2022年第4期1卷 11-21页

作者： Zhongjun Jiang yingjian liu Liang Wang Department of Optics and Optical Engineering University of Science and Technology of ChinaHefei 230026China.

Optical antennas play an important role in optical field *** them,nanoscale bowtie antennas have been extensively studied for its high confinement and *** this mini-review,we start with a brief introduction of bowtie antennas and underlying *** we review the applications with respect to optically and electrically excited nanoscale bowtie *** driven bowtie antennas enable a set of optical applications such as near-field imaging/trapping,nonlinear response,nanolithography,photon generation and ***,we put emphasis on the principle and applications of electrically driven bowtie antennas,an emerging method of generating ultrafast and broadband tunable *** a word,nanoscale bowtie antennas still have great potential research value to explore.

关键词： bowtie antenna near-field imaging nanolithography nonlinear nanolaser inelastic tunneling

Size-refinement enhanced flexibility and electrochemical performance of MXene electrodes for flexible waterproof supercapacitors

在线全文

学校读者我要写书评

暂无评论

Journal of Energy Chemistry 2021年第12期30卷 594-603,I0014页

作者： Jinkun Sun yingjian liu Jiayi Huang Jiatian Li Mengmeng Chen Xiaoyu Hu Yatao liu Run Wang Yanan Shen Jingjing Li Xuecheng Chen Dong Qian Baigang An Zunfeng liu Key Laboratory of Functional Polymer Materials College of Chemistry and College of PharmacyState Key Laboratory of Medicinal Chemical BiologyFrontiers Science Center for New Organic MatterNankai UniversityTianjin 300071China School of Chemical Engineering University of Science and Technology LiaoningAnshan 114051LiaoningChina Department of Mechanical Engineering The University of Texas at DallasRichardsonTX 75080United States Nanomaterials Physicochemistry Department Faculty of Chemical Technology and EngineeringWest Pomeranian University of TechnologySzczecinPiastów Ave.4271-065 SzczecinPoland

Increasing mechanical flexibility without sacrificing electrochemical performance of the electrode material is highly desired in the design of flexible electrochemical energy storage *** metal-related materials science,decreasing the grain size introduces more grain boundaries;this stops dislocations and crack propagation under deformation,and results in increased strength and ***,such a size refinement effect has not been considered in the mechanical properties,particle stacking,wetting,and electrochemical performances of flexible supercapacitor *** this paper,MXene was used as an electrode material to study the size refinement effect of flexible *** refinement improved the strength and toughness of the MXene electrodes,and this resulted in increased *** elemental analysis provided a theoretical understanding of size refinement-increased ***,the size refinement also improved the specific surface area,electric conductance,ion transportation,and water wetting properties of the electrode,and the size refinement provided highly increased energy density and power density of the MXene supercapacitors.A highly flexible,water-proof supercapacitor was fabricated using size-refined *** current study provides a new viewpoint for designing tough and flexible energy storage *** size refinement effect may also be applicable for metal ion batteries and electronic and photo devices composed of MXene and other nanoparticles.

关键词： Grain refinement Flexible supercapacitor Size grading MXene

Reduced Order Machine Learning Finite Element Methods:Concept,Implementation,and Future Applications Dedicated to Professor Karl Stark Pister for his 95th birthday

在线全文

学校读者我要写书评

暂无评论

Computer Modeling in Engineering & Sciences 2021年第12期129卷 1351-1371页

作者： Ye Lu Hengyang Li Sourav Saha Satyajit Mojumder Abdullah Al Amin Derick Suarez yingjian liu Dong Qian Wing Kam liu Department of Mechanical Engineering Northwestern UniversityEvanston60208USA Theoretical and Applied Mechanics Northwestern UniversityEvanston60208USA Department of Mechanical Engineering The University of Texas at DallasRichardson75080USA

This paper presents the concept of reduced order machine learning finite element(FE)*** particular,we propose an example of such method,the proper generalized decomposition(PGD)reduced hierarchical deeplearning neural networks(HiDeNN),called *** described first the HiDeNN interface seamlessly with the current commercial and open source FE *** proposed reduced order method can reduce significantly the degrees of freedom for machine learning and physics based modeling and is able to deal with high dimensional *** method is found more accurate than conventional finite element methods with a small portion of degrees of *** potential applications of the method,including topology optimization,multi-scale and multi-physics material modeling,and additive manufacturing,will be discussed in the paper.

关键词： Machine learning model reduction HiDeNN-PGD topology optimization multi-scale modeling additive manufacturing

Adaptive Regulator for Natural Gas Transportation based on DDPG

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

Adaptive Regulator for Natural Gas Transportation based on D...

瑞格列奈联合甘精胰岛素治疗2型糖尿病效果观察

第43届中国控制会议

作者： yingjian liu Xiaodong Zhang China University of Petroleum (East China)

In this paper,we propose an adaptive regulator for natural gas transportation based on Deep Deterministic Policy Gradient(DDPG).Currently,most natural gas pipeline pressure and flow control rely on traditional control ***,the transportation issues of natural gas pipelines are subject to frequent environmental *** instance,variations in factors such as temperature and density of natural gas can necessitate cumbersome readjustment and design of traditional *** the DDPG controller can automatically learn and optimize the control strategy during the training process by setting appropriate performance *** makes the DDPG controller more convenient and flexible,suitable for a wide range of control *** paper constructs a model of the natural gas pipeline environment,designs an appropriate control scheme,and simultaneously develops a DDPG controller,along with relevant performance *** the established natural gas pipeline model,simulation experiments verify the effectiveness of the DDPG control algorithm.

关键词： Adaptive regulator DDPG controller Pressure and flow control Reinforcement learning

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

河南医学研究 2017年第4期26卷 604-606页

作者：刘迎见商丘市第一人民医院内分泌科河南商丘476100

目的探究瑞格列奈联合甘精胰岛素治疗2型糖尿病的效果。方法选取于2014年7月至2015年5月在商丘市第一人民医院治疗的2型糖尿病患者60例作为研究对象,随机分为两组,各30例。对照组给予患者皮下注射甘精胰岛素治疗,观察组给予患者瑞格列... 详细信息

目的探究瑞格列奈联合甘精胰岛素治疗2型糖尿病的效果。方法选取于2014年7月至2015年5月在商丘市第一人民医院治疗的2型糖尿病患者60例作为研究对象,随机分为两组,各30例。对照组给予患者皮下注射甘精胰岛素治疗,观察组给予患者瑞格列奈联合皮下注射甘精胰岛素进行治疗,比较治疗后两组患者的血糖和Hb A1c水平、胰岛素用量、血糖达标时间、胰岛β细胞功能改善情况。结果观察组患者的血糖和Hb A1c水平、胰岛素用量、血糖达标时间、胰岛β细胞改善情况均优于对照组,差异有统计学意义(P<0.05)。结论实施瑞格列奈联合甘精胰岛素治疗2型糖尿病疗效显著,有利于预后。

关键词：瑞格列奈甘精胰岛素 2型糖尿病疗效

前列地尔联合厄贝沙坦对早期糖尿病肾病患者肾功能的影响

在线全文

学校读者我要写书评

暂无评论

临床研究 2015年第5期23卷 7-8页

作者：刘迎见河南省商丘市第一人民医院内分泌科 476100

目的探究前列地尔联合厄贝沙坦对早期糖尿病肾病患者肾功能的影响.方法随机选择2013年3月-2015年3月我院收治的早期糖尿病肾病患者60例为研究对象,依入院顺序分为两组,各30例.对照组在常规治疗基础上使用厄贝沙坦,观察组在对照组基础... 详细信息

目的探究前列地尔联合厄贝沙坦对早期糖尿病肾病患者肾功能的影响.方法随机选择2013年3月-2015年3月我院收治的早期糖尿病肾病患者60例为研究对象,依入院顺序分为两组,各30例.对照组在常规治疗基础上使用厄贝沙坦,观察组在对照组基础上联合使用前列地尔.将两组患者治疗后肾功能、治疗效果及生活质量进行对比.结果治疗后观察组SCr、Alb、BUN、UAER肾功能指标均明显优于对照组,差异有统计学意义(P<0.05);观察组治疗总有效显高于对照组,差异有统计学意义(P<0.05);观察组治疗后生活质量评分明显高于对照组,差异有统计学意义(P<0.05);治疗后观察组生活质量各维度评分均明显高于对照组,差异有统计学意义(P<0.05).结论前列地尔联合厄贝沙坦治疗早期糖尿病肾病患者可明显减少蛋白尿,优化肾功能,效果显著.

关键词：糖尿病肾病前列地尔厄贝沙坦肾功能

基于Edge Boxes和深度学习的非限制条件下人脸检测

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

现代电子技术 2018年第13期41卷 29-33页

作者：刘英剑张起贵太原理工大学数字图像处理实验室山西晋中030600

针对光线、旋转、遮挡、平移等因素对人脸检测结果产生的干扰,提出一种基于Edge Boxes和深度学习相结合的人脸检测算法。首先采用Edge Boxes算法提取出可能存在人脸的边界框,提取边界框中的图像并调整至合适的大小,作为卷积神经网络的输... 详细信息

针对光线、旋转、遮挡、平移等因素对人脸检测结果产生的干扰,提出一种基于Edge Boxes和深度学习相结合的人脸检测算法。首先采用Edge Boxes算法提取出可能存在人脸的边界框,提取边界框中的图像并调整至合适的大小,作为卷积神经网络的输入,然后利用卷积神经网络对提取出的图像进行特征提取和分类,最后利用非极大抑制算法排除多余人脸检测框,得到人脸的准确位置。该算法应用于LFW和Yale B人脸数据库的检测率分别达到98.7%和98.5%,识别单张人脸的时间均小于0.5 s。实验结果表明,该算法在检测率和检测速率方面较传统算法都有了很大的提高,对于遮挡、光照、旋转等干扰具有更强的鲁棒性。

关键词：人脸检测特征提取深度学习 Edge Boxes 卷积神经网络非极大抑制算法