检索结果-南通市图书馆

维普期刊数据库

Alterations of attention and impulsivity in the rat following a transgenerational decrease in dietary omega-3 fatty acids

在线全文

学校读者我要写书评

暂无评论

Food Science and Human Wellness 2018年第1期7卷 49-56页

作者： Joachim Hauser Ewelina Stollberg Andreas Reissmann Ivo Kaunzinger Klaus W.Lange Department of Experimental Psychology University of Regensburg93040 RegensburgGermany

Polyunsaturated fatty acids(PUFAs),particularly the omega-3 PUFAs,are thought to be involved in neuronal processes,to play a role in neurodevelopmental disorders and to be important for the integrity of central nervous system *** present study investigated the effects of nutritional omega-3 PUFAs on attentional functions and impulsive behavior in Wistar *** this purpose,female Wistar rats were fed an omega-3 deficient diet over several generations,and the dams of the seventh generation were randomly assigned to two diet groups and fed an omega-3 deficient or an omega-3 sufficient *** addition,a group of previously untreated dams was fed an omega-3 sufficient *** male offspring of these three diet groups were tested using an established paradigm for the assessment of attention and impulsive behavior,i.e.a modified version of the five-choice-serial-reaction-time task(5CSRTT).The present data show that the deficiency of omega-3 PUFAs over generations led to substantial changes in attentional processes and impulsive *** impairments associated with an omega-3 deficiency were partly corrected by treatment with the omega-3 sufficient diet in the last generation of the omega-3 deficient group which showed substantial improvements in attention *** there were no significant effects of dietary modifications on psychomotor activity levels,there was some evidence for changes in impulsive *** conclusion,transgenerational dietary changes in the availability of omega-3 PUFAs led to changes in attentional processes and impulsive behavior in rats,supporting the hypothesis that omega-3 PUFAs play a role in cognitive and behavioral *** present findings offer a promising approach in the investigation of the role of omega-3 PUFAs in a variety of cognitive and behavioral domains.

关键词： Omega-3 fatty acid Polyunsaturated fatty acids attention Impulsivity Activity Rat Neurodevelopmental disorders Psychiatric disorders

维普期刊数据库

Dual-energy CT Imaging via an Augmented attention Deep Learning Approach

在线全文

学校读者我要写书评

暂无评论

Dual-energy CT Imaging via an Augmented Attention Deep Learn...

Bi-attention:面向终端的细粒度识别网络加速方法

第十七届中国体视学与图像分析学术会议

作者：张文文 WAN Luping 孙保华赵维 Department of Medical Physics School of Physics at Beihang University

Compared to conventional single-energy CT(SECT),dual-energy CT(DECT) allows differentiation and quantification of different tissues and can be used for a wide spectrum of emerging clinical *** providing innumerable benefits,DECT has not been widely used yet,especially in less-developed region,mainly because of its ***,some deep learning-based approaches have been devoted to perform DECT imaging from easily available *** we propose an augmented attention U-Net which includes weight map generator and DECT imaging module to generate DECT images using standard SECT *** weight map generator provides pixel-level weights according to the features extracted from *** imaging module encompasses a classical U-Net and several augmented attention *** weight maps as input,the augmented attention gates can assist U-Net to focus on region of interest(ROI) to reduce the inference of CT noise in non-denoised *** performance of our approach was studied using images from patients who received contrast-enhanced abdomen DECT scan with popular DECT applications:virtual non-contrast(VNC) imaging and contrast *** relevant metrics were used for quantitative *** absolute HU difference between the original high-energy CT(HECT) images generated from DECT scan and the predicted ones by our method are2.6 HU,2.4 HU,2.0 HU,0.4 HU and 2.9 HU for the ROIs on heart,aorta,spine,liver and kidney,*** addition,comparing our method with the state-of-the-art deep learning-based DECT imaging method,we find that the distribution of ROI’s CT value predicted by our method are closer to ground truth,showing the merit of the proposed *** premise of the approach is that DECT can be attained without any additional measurement than a SECT *** technique provides a simple and costeffective solution for modern DECT clinical applications.

关键词： Dual-energy CT Single-energy CT Deep learning attention

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

高技术通讯 2023年第2期33卷 156-166页

作者：钟巧灵汪啸张志斌李冰程学旗中国科学院计算技术研究所网络数据科学与技术重点实验室北京100190 中国科学院大学计算机科学与技术学院北京100049

细粒度识别是针对具有微小差异的对象进行分类的图像识别任务,深度学习模型在细粒度识别任务上取得了较大的进步。然而现有的细粒度识别深度神经网络模型采用多个模型结构叠加,无法在手机、无人机等资源受限终端设备上部署。本文提出Bi-... 详细信息

细粒度识别是针对具有微小差异的对象进行分类的图像识别任务,深度学习模型在细粒度识别任务上取得了较大的进步。然而现有的细粒度识别深度神经网络模型采用多个模型结构叠加,无法在手机、无人机等资源受限终端设备上部署。本文提出Bi-attention细粒度识别模型加速方法,使用高效的TensorSketch运算以及权重共享机制,在Stanford Cars数据集上的准确率为91.6%,且比现有最先进的模型提高1.2%。本文提出一种结构化剪枝训练方法,通过LASSO正则化算法,在模型训练过程删除批归一化(BN)操作中不重要的扩展因子。实验结果表明,该剪枝方法可以降低Bi-attention模型大小为原来的1/4。

关键词：细粒度识别 attention 结构化剪枝 L1正则化终端

维普期刊数据库

Subjective experience and visual attention to a historic building:A real-world eyetracking study

在线全文

学校读者我要写书评

暂无评论

Frontiers of Architectural Research 2020年第4期9卷 774-804页

作者： Luis Alfonso de la Fuente Suarez Architecture School Universidad Autonoma de Nuevo LeonSon Nicolos de los Garza(Monterrey)66455Mexico

The objective of this exploratory study is to discover the relations between visual attention to an architectural work and the subjective experiences produced during its observation.A subjective experience with a building is the specific manner in which an architectural element or the building as a whole are presented to a *** qualities of the element as how they are perceived and the inner world of the observer are both considered in this subjective *** aims of this study are to describe the experience that a building generates in people,pinpoint what has attracted the participant's attention during each view of the building in an itinerary;and understand which aspects of this selected case study have made it an object of attend *** eye-tracking study was carried out using a portable eye *** tool allowed the participants to walk freely around the exterior of a historic train station in the city of Monterrey,*** groups of participants,with routes starting from opposite sides of the building,contemplated it while using the eye *** tracking allowed for the identification of the architectural elements that were objects of attention,the lengths of observation,and the points of view from which the elements were *** data collection techniques,such as the think-aloud protocol and a special type of survey,were used to deeply understand the experiences that accompanied the visual exploration of the historic *** suggest that the participants observed the different architectural elements for a time that was neither influenced by the route used to explore the building nor the point where that route was *** architectural elements identified as examples of high-quality architecture and perceived as aesthetically pleasing by the participants during the itinerary were observed for longer times.

关键词： Eye tracking Architecture attention Subjective experience Data visualization

引入位置信息和attention机制的诈骗电话文本分类

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

小型微型计算机系统 2023年第11期44卷 2502-2509页

作者：周俊杰许鸿奎卢江坤张子枫李振业郭文涛山东建筑大学信息与电气工程学院济南250000 山东建筑大学山东省智能建筑技术重点实验室济南250000

诈骗电话案件频频发生并威胁着人们的生活,含有诈骗语义的语句与内容的前后文、语句序列、局部相关内容以及显著关键词语相关.本文提出神经网络模型PEAGCNN(Position Embedding and attention are introduced into BiGRU and CNN)对诈... 详细信息

诈骗电话案件频频发生并威胁着人们的生活,含有诈骗语义的语句与内容的前后文、语句序列、局部相关内容以及显著关键词语相关.本文提出神经网络模型PEAGCNN(Position Embedding and attention are introduced into BiGRU and CNN)对诈骗电话文本分类.首先构建相关数据集,词嵌入用于表示文本,不同频率的正弦、余弦函数对文本位置信息编码并融入词嵌入向量,然后分别利用BiGRU(Bidirectional Gated Recurrent Unit)和CNN(Convolutional Neural Network)提取文本上下文相关信息、语句序列以及局部相关性,attention机制对提取出的信息重新分配权重,突出关键信息的作用,最后将两种信息融合,通过Softmax实现分类.在THUCNews数据集以及诈骗电话文本数据集上的实验结果表明,本文提出模型的准确率和F1值均比对比模型有提升,同时模型对诈骗电话文本数据集分类的各项性能指标均在0.91以上.

关键词：诈骗电话文本分类双向门控循环神经网络卷积神经网络位置编码 attention

维普期刊数据库博看期刊

Out-of-sync in managerial attention:Temporal and repertory mismatches between the headquarters and subsidiary

在线全文

学校读者我要写书评

暂无评论

International Journal of Innovation Studies 2019年第2期3卷 40-53页

作者： Armi Temmes Liisa Valikangas Department of Management Studies School of BusinessAalto UniversityP.O.Box 21230FIN-00076 AALTOFinland Department of Management and Organization Hanken School of EconomicsP.O.Box 47900101 HelsinkiFinland

We advance the attention-based view by presenting empirical evidence that the attention of headquarters and subsidiary managers in a multi-business organization is,at times,out of *** on empirical material that allows us to differentiate between what is attended to and what is ignored by management,we analyze the focus of managerial attention,environmental and subsidiary stimuli,and actions taken in the decision-making process over 15 years,during a period of strategic *** suggest that attentional mismatches occur not only between strategic issues but also between what are considered relevant responses or actions to be taken at any particular *** analyze the origins of the attentional mismatches and explore ways to avoid such nonalignment in strategic decision making.

关键词： attention attention-based view Decision making Headquarters Strategy Subsidiary

融合attention的BiLSTM＿CRF藏文分词技术研究

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

融合Attention的BiLSTM＿CRF藏文分词技术研究

作者：常芳玉青海师范大学

学位级别：硕士

随着互联网的普及,文本信息数量急剧增长,需要更加精确、高效地进行校对、分类、翻译以及信息检索等自然语言处理任务。词作为表达语义信息的最小单位,使得切分得到的词序列具有语义信息。为了便于计算机能够理解人类的语言,以便更好地... 详细信息

随着互联网的普及,文本信息数量急剧增长,需要更加精确、高效地进行校对、分类、翻译以及信息检索等自然语言处理任务。词作为表达语义信息的最小单位,使得切分得到的词序列具有语义信息。为了便于计算机能够理解人类的语言,以便更好地提取出其中的重要内容,我们需要对文本进行分词,因此分词是自然语言处理的一项基础性工作。本文以词位标签集的设计和分词模型的设计及实现等方面研究了藏文分词技术。主要内容包括:(1)提出了八词位标签标注方法为了学习到更多的语义信息,从而提升分词的性能,本文在四词位标签集的基础上,通过扩充得到了一个八词位标签集,并制定了八词位标签集标注规则。基于词位标注的藏文分词需要大量的实验语料,使用人工对语料进行标注,不仅效率低且容易出错,为此设计了八词位标签集标注算法。(2)设计了融合attention的BiLSTM＿CRF藏文分词模型融合BiLSTM和CRF的BiLSTM＿CRF虽然能够自动获取上下文信息,且能考虑输出标签间的关系,但是BiLSTM＿CRF不能突出局部重点信息,并且丢失长序列历史特征。针对存在的问题,本文在BiLSTM＿CRF基础上,提出了一种融合attention的BiLSTM＿CRF藏文分词模型attention＿BiLSTM＿CRF。模型首先通过BiLSTM获取全局上下文信息,再利用attention加强局部重点信息,并缓解了历史信息丢失造成的影响,最后利用CRF学习标签间的关系来规避不合法标签。(3)实验验证了八词位标签标注方法和attention＿BiLSTM＿CRF藏文分词模型的有效性为了验证八词位标签标注方法和attention＿BiLSTM＿CRF藏文分词模型的有效性,本文分别在四词位标签集、六词位标签集以及八词位标签集下,使用CRF、BiLSTM、BiLSTM＿CRF以及attention＿BiLSTM＿CRF等藏文分词模型进行了实验。实验表明,八词位标签集下当attention类型为Spare Self,嵌入在正向LSTM和反向LSTM后时模型attention＿BiLSTM＿CRF分词性能最优。

关键词：自然语言处理藏文分词 attention BiLSTM CRF

Effect of classroom-based physical activity interventions on attention and on-task behavior in schoolchildren:A systematic review

同方学位论文库评论

在线全文

同方学位论文库

学校读者我要写书评

暂无评论

Sports Medicine and Health Science 2021年第3期3卷 125-133页

作者： Stefanie Ruhland Klaus W.Lange Institute of Psychology University of RegensburgGermany

Physical activity in children is associated with several cognitive *** children and adolescents spend an increasing amount of time engaged in sedentary behavior both at school and in their free time,movement breaks during class hours,in which students are physically active,may be beneficial for effective *** aim of this systematic research is to provide an overview of prospective studies investigating the influence of classroom-based physical activity(CB-PA)interventions on attention and on-task behavior in school-aged children and adolescents aged between 4 and 18 years.A systematic search of electronic databases(PubMed,Science Direct,PsycINFO,Ovid),according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses(PRISMA)statement,was performed from July 2020 to March *** characteristics data were analyzed and a methodological quality assessment,using a modified Downs and Black checklist,of both randomized and non-randomized studies was ***,the available evidence points to a beneficial effect of exercise on attention and on-task behavior in a classroom ***,methodological differences concerning partici-pants and duration and type of physical activity should be considered when comparing the *** studies with more comparable methodology are needed to provide a better understanding of the effect of CB-PA on attention and on-task behavior.

关键词： attention Children Classroom On-task behavior Physical activity

Image understanding, attention and human early visual cortex

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Frontiers of Electrical and Electronic Engineering in China 2012年第1期7卷 85-93页

作者： Fang FANG Yizhou WANG Department of Psychology Peking University Beijing 100871 China Key Laboratory of Machine Perception （Ministry of Education） Peking University Beijing 100871 China Center for Life Sciences Peking University Beijing 100871 China National Engineering Lab for Video Technology School of Electronics Engineering and Computer Science Peking University Beijing 100871 China

This paper reviews our recent fMRI and psychophysical finding on： 1） perceived size represen- tation in V1; 2） border ownership representation in V2; and 3） neural processing of partially occluded face. These findings demonstrate that the human early vi- sual cortex not only performs local feature analyses, but also contributes significantly to high-level visual computation with assistance of attention-enabled cortical feed- back. Moreover, by taking advantage of recent findings on early visual cortex from neuroscience and cognitive science, we build a biologically plausible attention model that can well predict human scanpaths on natural images.

关键词： vision attention image understanding early visual cortex