基于Contextual Transformer的自动驾驶单目3D目标检测

针对当前单目3D目标检测中存在的漏检和多尺度目标检测效果不佳的问题,提出了一种基于Contextual Transformer的自动驾驶单目3D目标检测算法(CM-RTM3D)。在ResNet-50网络中引入Contextual Transformer(CoT),构建ResNet-Transformer架构以提取特征。设计多尺度空间感知模块(MSP),通过尺度空间响应操作改善浅层特征的丢失情况,嵌入沿水平和竖直两个空间方向的坐标注意力机制(CA),使用softmax函数生成各尺度的重要性软权重。在偏移损失中采用Huber损失函数代替L1损失函数。实验结果表明:在KITTI自动驾驶数据集上,相较于RTM3D算法,该算法在简单、中等、困难三个难度级别下,AP3D分别提升了4.84、3.82、5.36个百分点,APBEV分别提升了4.75、6.26、3.56个百分点。

融合MobileNet与Contextual Transformer的人脸识别研究

FaceNet作为人脸识别的一大跨越,以其高精度、低硬件配置等优势被广泛应用于各个人脸识别相关领域。本文开源了首个餐厅支付场景下的中国人脸数据集CN-Face,该数据集拥有13000人的人脸图像,总计100000张。此外,本文以CA-SIA-WebFace作为训练集,利用改进后的MobileNet主干网络,采取不同的注意力机制添加方法,改变激活函数并且融入Contextual Transformer模块,大大降低了参数量和识别速度,显著提升了人脸识别精度。相较于原版FaceNet,在LFW测试集下,准确率达到98.79%,提升了2.74%,在CN-Face数据集中准确率达到95.22%,提升了1.35%。

Resilience, the 6th Vital Sign: Conceptualizing, Contextualizing, and Operationalizing All Six Vital Signs

There are five vital signs that healthcare providers assess: temperature, pulse, respiration, blood pressure, and pain. Normal levels for the five vital signs are published by the American Heart Association, and other specialty organizations, however, the sixth vital sign (resilience) which adopts the measure of immune resilience is suggested in this paper. Resilience is the ability of the immune system to respond to attacks and defend effectively against infections and inflammatory stressors, and psychological resilience is the capacity to resist, adapt, recover, thrive, and grow from a challenge or a stressor. Individuals with better optimal immune resilience had better health outcomes than those with minimal immune resilience. The purpose of this paper is to conceptualize, contextualize, and operationalize all six vital signs. We suggest measuring resilience subjectively and objectively. Subjectively, use a 5-item guided interview revised from the Connor-Davidson Resilience Scale (CDRC), a scale of 10 items. The revised CDRC scale is a 5-item scale. The scale is rated on a 5-point Likert scale from 0 (not true) to 4 (true all the time). The total score ranges from 0 to 20, with higher total scores indicating greater resilience. The scale demonstrated good construct validity and internal consistency (α = 0.85) during the development of the scale. The CD-RISC had a good Cronbach’s alpha level of 0.85. The Revised CD-RISC can be completed in 2 - 4 minutes. To measure resilience objectively, we suggest using Immune Resilience (IR) levels, the level of resilience to preserve and/or rapidly restore immune resilience functions that promote disease resistance and control inflammation and other inflammatory stress. IR levels are gauged with two peripheral blood metrics that quantify the balance between CD8 and CD4 T-cell levels and gene expression signatures tracking longevity-associated immunocompetence and mortality- or entropy-associated inflammation. IR deregulation is potentially reversible by decreasing inflammatory stress. IR metrics and mechanisms have utility as vital signs and biomarkers for measuring immune health and improving health outcomes.

The Impact of Contextual Teaching Method on Sichuan Folk Song Education and Students’Musical Expressiveness

This study explores the application of the contextual teaching method in Sichuan folk song education and its impact on students’musical expressiveness.By incorporating contextual teaching methods in music classes,this research investigates the effectiveness of this approach in enhancing students’understanding of Sichuan folk songs and improving their musical expressiveness and emotional expression.A mixed-method research approach is employed,utilizing classroom observations,questionnaires,interviews,and statistical analysis to assess the practical outcomes of contextual teaching in folk song education.

Research on the Influence of Anchor Attributes on Consumers’Online Behaviors in Social E-Commerce Platforms:The Moderating Effect of Platform Contextual Factors

As e-commerce continues to mature,the advantages of live streaming within the industry have become increasingly apparent,offering significant growth opportunities.Social e-commerce platforms,which are user-centered,integrate social networks with e-commerce by leveraging social interactions to drive product sales and enhance the overall consumer shopping experience.This type of e-commerce fosters engagement and promotes products by merging online communities with shopping behavior,creating a more interactive and dynamic marketplace.It not only retains the traditional e-commerce trading and marketing functions but also adds a social dimension,making live stream anchors crucial figures connecting consumers with products.These anchors can attract consumers with their appearance and charm,and use their expertise on live streaming platforms to guide consumers by recommending live content.They can also interact with their audiences and potentially influence them to purchase the recommended goods.It is evident that the attributes of anchors in live streaming rooms significantly impact consumers’online behavior.Therefore,researching how platform contextual factors regulate consumers’online behavior is of great practical significance.This study employs multilevel regression analysis to support its hypotheses using data.The findings indicate that contextual factors of the platform significantly influence online behavior,enhancing the positive relationship between user attachment and online activities.

Facial expression recognition with contextualized histograms

A new algorithm taking the spatial context of local features into account by utilizing contextualized histograms was proposed to recognize facial expression. The contextualized histograms were extracted fromtwo widely used descriptors—the local binary pattern（ LBP） and weber local descriptor（ WLD）. The LBP and WLD feature histograms were extracted separately fromeach facial image,and contextualized histogram was generated as feature vectors to feed the classifier. In addition,the human face was divided into sub-blocks and each sub-block was assigned different weights by their different contributions to the intensity of facial expressions to improve the recognition rate. With the support vector machine（SVM） as classifier,the experimental results on the 2D texture images fromthe 3D-BU FE dataset indicated that contextualized histograms improved facial expression recognition performance when local features were employed.

On the effects of contextualized explanation and Ebbinghaus Forgetting Curve on the teaching and learning of English vocabulary

This paper reports the outcomes of three vocabulary tests taken by 71 second-year undergraduates, discusses the possible effects of contextualized explanation of new words and Ebbinghaus Forgetting Curve on the vocabulary teaching and learning. The authors find that in a short duration there is a significant difference between the effect of bilingual （English ＆ Chinese） explanation and that of monolingual （Chinese） explanation on the students＇ recognition of English new words.

Short Video Recommendation Algorithm Incorporating Temporal Contextual Information and User Context

With the popularity of 5G and the rapid development of mobile terminals,an endless stream of short video software exists.Browsing short-form mobile video in fragmented time has become the mainstream of user’s life.Hence,designing an efficient short video recommendation method has become important for major network platforms to attract users and satisfy their requirements.Nevertheless,the explosive growth of data leads to the low efficiency of the algorithm,which fails to distill users’points of interest on one hand effectively.On the other hand,integrating user preferences and the content of items urgently intensify the requirements for platform recommendation.In this paper,we propose a collaborative filtering algorithm,integrating time context information and user context,which pours attention into expanding and discovering user interest.In the first place,we introduce the temporal context information into the typical collaborative filtering algorithm,and leverage the popularity penalty function to weight the similarity between recommended short videos and the historical short videos.There remains one more point.We also introduce the user situation into the traditional collaborative filtering recommendation algorithm,considering the context information of users in the generation recommendation stage,and weight the recommended short-formvideos of candidates.At last,a diverse approach is used to generate a Top-K recommendation list for users.And through a case study,we illustrate the accuracy and diversity of the proposed method.

Aspect-Based Sentiment Classification Using Deep Learning and Hybrid of Word Embedding and Contextual Position

Aspect-based sentiment analysis aims to detect and classify the sentiment polarities as negative,positive,or neutral while associating them with their identified aspects from the corresponding context.In this regard,prior methodologies widely utilize either word embedding or tree-based rep-resentations.Meanwhile,the separate use of those deep features such as word embedding and tree-based dependencies has become a significant cause of information loss.Generally,word embedding preserves the syntactic and semantic relations between a couple of terms lying in a sentence.Besides,the tree-based structure conserves the grammatical and logical dependencies of context.In addition,the sentence-oriented word position describes a critical factor that influences the contextual information of a targeted sentence.Therefore,knowledge of the position-oriented information of words in a sentence has been considered significant.In this study,we propose to use word embedding,tree-based representation,and contextual position information in combination to evaluate whether their combination will improve the result’s effectiveness or not.In the meantime,their joint utilization enhances the accurate identification and extraction of targeted aspect terms,which also influences their classification process.In this research paper,we propose a method named Attention Based Multi-Channel Convolutional Neural Net-work(Att-MC-CNN)that jointly utilizes these three deep features such as word embedding with tree-based structure and contextual position informa-tion.These three parameters deliver to Multi-Channel Convolutional Neural Network(MC-CNN)that identifies and extracts the potential terms and classifies their polarities.In addition,these terms have been further filtered with the attention mechanism,which determines the most significant words.The empirical analysis proves the proposed approach’s effectiveness compared to existing techniques when evaluated on standard datasets.The experimental results represent our approach outperforms in the F1 measure with an overall achievement of 94%in identifying aspects and 92%in the task of sentiment classification.

基于CRV-YOLO的苹果中心花和边花识别方法

苹果树疏花是果园生产管理中的重要环节。准确高效地识别苹果中心花和边花,是研发智能疏花机器人的前提。针对苹果疏花作业中的实际需求,提出了一种基于CRV-YOLO的苹果中心花和边花识别方法。本文基于YOLO v5s模型进行了如下改进:将C-CoTCSP结构融入Backbone,更好地学习上下文信息并提高了模型特征提取能力,提高了模型对外形相似和位置关系不明显的中心花和边花的检测性能。在Backbone中添加改进RFB结构,扩大特征提取感受野并对分支贡献度进行加权,更好地利用了不同尺度特征。采用VariFocal Loss损失函数,提高了模型对遮挡等场景下难识别样本检测能力。在3个品种1837幅图像数据集上进行了实验,结果表明,CRV-YOLO的精确率、召回率和平均精度均值分别为95.6%、92.9%和96.9%,与原模型相比,分别提高3.7、4.3、3.9个百分点,模型受光照变化和苹果品种影响较小。与Faster R-CNN、SSD、YOLOX、YOLO v7模型相比,CRV-YOLO的精确率、平均精度均值、模型内存占用量和复杂度性能最优,召回率接近最优。研究成果可为苹果智能疏花提供技术支持。

A Personalized Adverse Drug Reaction Early Warning Method Based on Contextual Ontology and Rules Learning

Background: The fatality of adverse drug reactions (ADR) has become one of the major causes of the non-natural disease deaths globally, with the issue of drug safety emerging as a common topic of concern. Objective: The personalized ADR early warning method, based on contextual ontology and rule learning, proposed in this study aims to provide a reference method for personalized health and medical information services. Methods: First, the patient data is formalized, and the user contextual ontology is constructed, reflecting the characteristics of the patient population. The concept of ontology rule learning is then proposed, which is to mine the rules contained in the data set through machine learning to improve the efficiency and scientificity of ontology rule generation. Based on the contextual ontology of ADR, the high-level context information is identified and predicted by means of reasoning, so the occurrence of the specific adverse reaction in patients from different populations is extracted. Results: Finally, using diabetes drugs as an example, contextual information is identified and predicted through reasoning, to mine the occurrence of specific adverse reactions in different patient populations, and realize personalized medication decision-making and early warning of ADR.

体育与健康课程大单元教学中核心素养、结构化、情境化:要义阐释与关联表征

课程标准是体育与健康课程的顶层文本,是当下开展大单元教学的刚性要求。在课程标准的话语体系中核心素养、结构化、情境化是新理念的重要组成,成为基于课程标准的体育与健康课程大单元教学的重要支点。体育与健康课程大单元教学作为实现体育课堂教学改革的新型教学组织形式,也是核心素养、结构化、情境化落实的中介桥梁。教师唯有深刻认识核心素养、结构化、情境化的价值与内涵,把握彼此的相互关系,才能对体育与健康课程大单元教学有更明确的认识,从而有效指引专项运动技能教学的实践。核心素养统领大单元学习目标,贯穿于大单元教学始终,是学生全面发展的重要保障;结构化指引大单元学习内容,清晰呈现知识与技能的逻辑关系和层次结构,有助于学生更好地理解专项运动技能的内在逻辑和规律;情境化变革大单元学习活动,将抽象的知识和核心素养融入真实、具体的情境中,让学生在真实情境、实践共同体的互动中理解专项运动技能的实际价值和意义。体育与健康课程大单元教学中核心素养、结构化和情境化三者相互制约、相互促进,共同构成学生专项运动技能学习与全面发展的关键所在。

“中西交融”的管理理论建构与创新

基于理论与情境的匹配来建立管理理论分析框架,管理研究可以划分为成熟理论研究、拓展理论研究、新创理论研究和深入理论研究,分别解决不同的现实问题。中国管理理论的建构与创新主要延续着从“再情境化”到“去情境化”的动态循环,并在情境化理论和理论化情境的交替作用下,形成“成熟理论—拓展理论—新创理论—深入理论”的演进路径。中国情境不仅可以修正和完善西方管理理论并将其运用于本土情境,而且可以立足本土情境的独特性来建构中国管理理论,还能通过理论的持续创新进行扩散和传播,以形成能得到中西方共同接受与认可的管理理论体系,最终确立管理理论的“中国话语体系”和贡献中国管理智慧。

治理竞赛、创新扩散与情境不兼容:基层治理创新悬浮现象的发生逻辑——基于A市居委会轮岗制的案例研究

创新扩散和创新悬浮现象是基层治理研究的重要议题,但围绕两个议题的整合性研究并不多。基于对A市居委会轮岗制扩散过程和应用结果的案例研究,发现基层治理创新的构成要素和应用场景在扩散中发生明显变化。创新悬浮现象发生的起点是基层治理竞赛的评价标准具有模糊性,模糊性评价造就了在扩散中竞争和创新再生产等行动策略,并导致创新与创新使用者在工作目标、工作经验和工作激励等方面形成情境不兼容问题,进而使创新使用者采取名实分离策略。扩散中的创新再生产可以分为复制、加码、移植和转换四种,情境不兼容包括效率、竞争和社会三种情况,表明创新悬浮现象可以通过多种路径形成。扩散带来的创新悬浮问题表明作为创新调适机制的扩散可能存在积极扩散和消极扩散两种功能。破解基层治理创新悬浮问题有赖于营造更多的积极扩散,在此过程中需要平衡好创新标准化与基层自主性、局部创新与整体创新以及管理知识与执行知识等关系。

基于Transformer的多尺度遥感语义分割网络

为了提升遥感图像语义分割效果,本文针对分割目标类间方差小、类内方差大的特点,从全局上下文信息和多尺度语义特征2个关键点提出一种基于Transformer的多尺度遥感语义分割网络(muliti-scale Transformer network,MSTNet)。其由编码器和解码器2个部分组成,编码器包含基于Transformer改进的视觉注意网络(visual attention network,VAN)主干和基于空洞空间金字塔池化(atrous spatial pyramid pooling, ASPP)结构改进的多尺度语义特征提取模块(multi-scale semantic feature extraction module, MSFEM)。解码器采用轻量级多层感知器(multi-layer perception,MLP)配合编码器设计,充分分析所提取的包含全局上下文信息和多尺度表示的语义特征。MSTNet在2个高分辨率遥感语义分割数据集ISPRS Potsdam和LoveDA上进行验证,平均交并比(mIoU)分别达到79.50%和54.12%,平均F1-score(m F1)分别达到87.46%和69.34%,实验结果验证了本文所提方法有效提升了遥感图像语义分割的效果。

基于韵律的英语介词短语挂靠歧义消解研究

在言语交际过程中,说话人口语表达的韵律线索对交际双方的语言理解发挥着重要作用。为了深入了解中国英语学习者英语口语产出的韵律线索与预期表达含义之间的相关性,本文以中国大学英语学习者为研究对象,采用基于情境的半开放式口语产出任务,考察在特定的情境化实验条件下,被试对英语介词短语挂靠造成的歧义句的朗读任务完成情况。研究结果表明:中国大学英语学习者具有对口语韵律特征与句法结构之间关系的潜在敏感性,能够通过口语的韵律停顿和音高重音实现对英语介词短语动、名词挂靠引起的歧义的消解。本研究结果进一步证实了英语口语韵律特征与预期含义之间的相互关系,凸显了韵律训练在大学英语课堂教学中的重要性,为高校英语听说及阅读教学实践的开展提供了参考建议。

情境视角下儿童农耕研学服务体验设计研究--以钱唐农园为例

针对传统农耕研学体验单一、家庭参与度弱、社会关注度低等问题,从设计视角重构农耕研学创新服务体系,探讨基于情境体验的农耕研学服务体验设计研究。在情境体验理论与用户体验设计相关方法的指导下,研究儿童、家长对农园情境的理性和感性认知,挖掘其需求痛点与期望,归纳整理出农耕研学服务体验的三大情境需求。从情境体验设计的角度提出满足儿童农耕研学服务体验的设计策略并进行设计实践。乐学乐教、智能情感、可持续性的情境体验符合儿童农耕研学的需求,以设计实践验证了情境体验设计方法的应用价值。

奖励式众筹的后续市场效应:基于场景视角的研究

场景化赋予奖励式众筹独特的情感体验与社群价值,使得国内的众筹模式逐渐脱离早期单一的融资属性,向新品推广、品牌孵化、场景营销等多元化功能演变。文章选取小米有品众筹平台数据作为研究样本,基于场景视角探讨了奖励式众筹的场景化运作与价值创造逻辑,通过有序Logistic模型考察了奖励式众筹的后续市场效应,以及场景营销、场景互动的调节作用。研究发现,奖励式众筹绩效对于产品后续市场表现具有显著的正向影响,场景营销能够强化这一关系,而普通用户互动比领先用户互动的调节作用更为显著。异质性分析发现,奖励式众筹的后续市场效应在平台嵌入程度较高的企业中更为显著。进一步研究发现,奖励式众筹具有品牌推广效应,能够显著提升企业的品牌关注度。文章基于场景视角拓展了众筹研究的边界,对奖励式众筹平台及企业的场景化创新具有一定的指导意义。

管理研究的情境基础观:内涵、要素与框架

文章针对传统管理理论范式在解构复杂管理实践上存在的局部性与割裂性缺陷,提出并构建了情境基础观。首先,基于情境理论与传统管理理论范式,提出并解析了情境基础观的概念内涵;其次,解构情境基础观的要素结构,初步描画了由基本要素和独特要素构成的“情境谱系”;再次,通过多重情境要素关系的处理实现“情境深化”,并形成以情境谱系和情境深化相整合的情境化为内核,以“管理实践—情境化—管理理论”为外缘的情境基础观框架;最后,结合多元研究方法,提出理论与情境相协同的本土管理理论构建路径,以期为构建本土管理理论体系以及管理的中国学派作出新的探索。

面向视频数据的多模态情感分析

多模态情感分析旨在从文本、图像和音频数据中提取和整合语义信息,从而识别在线视频中说话者的情感状态。尽管多模态融合方案在此研究领域已取得一定成果,但是已有方法在处理模态间分布差异和关系知识的融合方面仍有欠缺,为此,提出一种多模态情感分析方法。设计一种多模态提示门(MPG)模块,其能够将非语言信息转换为融合文本上下文的提示,利用文本信息对非语言信号的噪声进行过滤,得到包含丰富语义信息的提示,以增强模态间的信息整合。此外,提出一种实例到标签的对比学习框架,在语义层面上区分隐空间中的不同标签以进一步优化模型输出。在3个大规模情感分析数据集上的实验结果表明,该方法的二分类精度相对次优模型提高了约0.7%,三分类精度提高了超过2.5%,达到0.671。该方法能够为将多模态情感分析引入用户画像、视频理解、AI面试等领域提供参考。