MSC-YOLO:Improved YOLOv7 Based on Multi-Scale Spatial Context for Small Object Detection in UAV-View

Accurately identifying small objects in high-resolution aerial images presents a complex and crucial task in thefield of small object detection on unmanned aerial vehicles(UAVs).This task is challenging due to variations inUAV flight altitude,differences in object scales,as well as factors like flight speed and motion blur.To enhancethe detection efficacy of small targets in drone aerial imagery,we propose an enhanced You Only Look Onceversion 7(YOLOv7)algorithm based on multi-scale spatial context.We build the MSC-YOLO model,whichincorporates an additional prediction head,denoted as P2,to improve adaptability for small objects.We replaceconventional downsampling with a Spatial-to-Depth Convolutional Combination(CSPDC)module to mitigatethe loss of intricate feature details related to small objects.Furthermore,we propose a Spatial Context Pyramidwith Multi-Scale Attention(SCPMA)module,which captures spatial and channel-dependent features of smalltargets acrossmultiple scales.This module enhances the perception of spatial contextual features and the utilizationof multiscale feature information.On the Visdrone2023 and UAVDT datasets,MSC-YOLO achieves remarkableresults,outperforming the baseline method YOLOv7 by 3.0%in terms ofmean average precision(mAP).The MSCYOLOalgorithm proposed in this paper has demonstrated satisfactory performance in detecting small targets inUAV aerial photography,providing strong support for practical applications.

A Web Service for Managing Spatial Context Dedicated to Serious Games on and for Smartphones

The GeoEduc3D project aims to provide educational games for smartphones based on Geomatics and use augmented reality techniques in order to make these games more immersive. To improve the immersive and interactive aspects of those games, we focused on the exploitation of spatial context in this particular application framework (serious games, augmented reality, smart phones, and multi-users environment). Our work has thus led to the design of a solution dedicated to the management of spatial context in a multi-players environment on and for smartphones. Several contributions have been made: modeling spatial context, proposing a service-oriented architecture to manage this context, defining a Web Service Spatial Context (WSCS) and implementation of a WSCS prototype and a mobile client according to an environment exploiting FourSquare, a geo-social application.

A formal model for access control with supporting spatial context

There is an emerging recognition of the importance of utilizing contextual information in authorization decisions. Controlling access to resources in the field of wireless and mobile networking require the definition of a formal model for access control with supporting spatial context. However, traditional RBAC model does not specify these spatial requirements. In this paper, we extend the existing RBAC model and propose the SC-RBAC model that utilizes spatial and location-based information in security policy definitions. The concept of spatial role is presented, and the role is assigned a logical location domain to specify the spatial boundary. Roles are activated based on the current physical position of the user which obtsined from a specific mobile terminal. We then extend SC-RBAC to deal with hierarchies, modeling permission, user and activation inheritance, and prove that the hierarchical spatial roles are capable of constructing a lattice which is a means for articulate multi-level security policy and more suitable to control the information flow security for safety-critical location-aware information systems. Next, con- strained SC-RBAC allows express various spatial separations of duty constraints, location-based cardinality and temporal constraints for specify fine-grained spatial semantics that are typical in location-aware systems. Finally, we introduce 9 in- variants for the constrained SC-RBAC and its basic security theorem is proven. The constrained SC-RBAC provides the foundation for applications in need of the constrained spatial context aware access control.

Encoding Spatial Context for Large-Scale Partial-Duplicate Web Image Retrieval

Many recent state-of-the-art image retrieval approaches are based on Bag-of-Visual-Words model and represent an image with a set of visual words by quantizing local SIFT（scale invariant feature transform） features. Feature quantization reduces the discriminative power of local features and unavoidably causes many false local matches between images, which degrades the retrieval accuracy. To filter those false matches, geometric context among visual words has been popularly explored for the verification of geometric consistency. However, existing studies with global or local geometric verification are either computationally expensive or achieve limited accuracy. To address this issue, in this paper, we focus on partialduplicate Web image retrieval, and propose a scheme to encode the spatial context for visual matching verification. An efficient affine enhancement scheme is proposed to refine the verification results. Experiments on partial-duplicate Web image search, using a database of one million images, demonstrate the effectiveness and efficiency of the proposed approach.Evaluation on a 10-million image database further reveals the scalability of our approach.

Mapping ideas from cyberspace to realspace: visualizing the spatial context of keywords from web page search results

We introduce a new method for visualizing and analyzing information landscapes of ideas and events posted on public web pages through customized web-search engines and keywords.This research integrates GIScience and web-search engines to track and analyze public web pages and their web contents with associated spatial relationships.Web pages searched by clusters of keywords were mapped with real-world coordinates(by geolocating their Internet Protocol addresses).The resulting maps represent web information landscapes consisting of hundreds of populated web pages searched by selected keywords.By creating a Spatial Web Automatic Reasoning and Mapping System prototype,researchers can visualize the spread of web pages associated with specific keywords,concepts,ideas,or news over time and space.These maps may reveal important spatial relationships and spatial context associated with selected keywords.This approach may provide a new research direction for geographers to study the diffusion of human thought and ideas.A better understanding of the spatial and temporal dynamics of the‘collective thinking of human beings’over the Internet may help us understand various innovation diffusion processes,human behaviors,and social movements around the world.

MFFNet:多级特征融合图像语义分割网络

在图像语义分割任务中,大多数方法未将不同尺度、不同层次的特征充分利用就直接进行上采样,会造成一些有效信息被当成冗余信息而被摒弃,从而降低对某些细小类别和相似类别分割的准确性和敏感性。为此,提出一个多级特征融合网络(MFFNet)。MFFNet采用编码器-解码器结构,在编码阶段,通过上下文信息提取路径和空间信息提取路径分别获取上下文信息与空间细节信息,增强像素间关联性与边界准确性;解码阶段设计一条多级特征融合路径,利用混合双边融合模块融合上下文信息;利用高低特征融合模块融合深层信息与空间信息;利用全局通道融合模块获取不同通道之间的联系,实现不同尺度信息的全局融合。MFFNet网络在PASCAL VOC 2012和Cityscapes验证集上的平均交互比(MIoU)分别为80.70%和76.33%,取得了较好的分割结果。

CIEFRNet:面向高速公路的抛洒物检测算法

高速公路抛洒物危及行车安全,极易诱发交通事故,及时识别并清理高速公路抛洒物十分重要。由于高速公路抛洒物在图像中面积占比小且图像背景复杂,现有检测方法常出现漏检和误检的情况。针对上述问题,提出了一种基于上下文信息增强和特征提纯的抛洒物检测算法,记为CIEFRNet。设计了一种融合上下文Transformer的主干特征提取模块(CSP-COT),充分挖掘局部静态上下文信息和全局动态上下文信息,增强小抛洒物的特征表示;主干网络中使用改进的空间金字塔池化(ISPP),通过级联的空洞卷积实现特征的多尺度下采样,减轻目标细节信息的损失;为提高特征融合能力,设计了特征提纯模块(CNAB),其中嵌入了提出的一种混合注意力机制(ECSA),可抑制图像背景噪声,强化微小抛洒物的特征;引入基于动态非单调聚焦机制的WIoU优化损失函数,提高小抛洒物学习能力,加速网络收敛。实验结果表明,所提方法在自制的高速公路抛洒物数据集上的精确率、召回率、AP0.5和AP0.5:0.95分别达到96.5%、81.6%、88.1%和46.5%,优于当前主流的目标检测方法,其算法复杂度也更低,满足实际场景应用需要。

宗族型传统聚落空间形态特征与保护研究——以山西省介休市南贾村为例

宗族型传统聚落往往基于单个或多个宗族聚居发展形成,宗族的兴衰变迁与行为活动影响着聚落本身的空间形态特征。文章对山西省介休市的宗族型传统聚落南贾村进行剖析,分析其宗族文脉与宗族结构,以及聚落空间形成演变中的不同阶段,解读南贾村的宗族结构与聚落空间形态特征的联系。并且根据研究和现场调研情况,对南贾村空间形态的现状进行总结,分析聚落空间保护中存在的问题和不足,基于宗族视角提出相应的保护策略方法,为宗族型传统聚落保护提供借鉴和参考。

新世纪以来中国数字乡村研究的发展脉络和展望

以2001-2023年CNKI收录的1404篇核心文献为样本,运用CiteSpace工具绘制知识图谱,对新世纪以来数字乡村理论文献进行可视化分析。研究发现:时间序列上,起始成长期稳定发展、高峰衰退期衔接过渡、再创辉煌期热度激增;空间分布上,经济学视角占据主导,科研横向合作与纵向批判继承发展存在短板,高频被引文献在整体框架建构与技术“赋能”层面发挥重要作用。研究主题上,主要围绕数字乡村战略背景和理论支撑、概念内涵和运行机制、实现路径和风险评估展开;重点领域数字化转型机制与数字乡村高质量发展的创新路径是研究热点。研究趋势上,“数字乡村”与共同富裕、城乡融合、现代化转型、数字中国建设等重大目标的衔接式研究值得重点关注。后续研究亟需在数字乡村未来形态、主体转型、整合机制及理论建构等方面深入推进。

基于双分支交互的实时语义分割算法

针对双分支实时语义分割算法存在双分支交互差、多尺度上下文信息提取不完善等问题,提出了基于双分支交互的实时语义分割网络(dual-branch interactive multi-scale fusion network for real-time semantic segmentation,DIMFNet)。算法以引导聚合双边语义分割网络(bilateral network with guided aggregation for real-time semantic segmentation,BiseNetV2)的双分支结构为基准进行改进,空间分支提取空间细节特征,上下文分支提取深层上下文特征。结合注意力思想提出注意力引导高级语义融合模块(attention guide high-level semantics fusion module,AGHSM)实现双分支的交互融合,以获得更好的空间特征表示;对金字塔池化模块进行改进,提出采用多层聚合金字塔池化模块(multi-layer aggregation pyramid pooling module,MAPPM)提取多尺度上下文特征,以获得更好的上下文特征表示。算法在Cityscapes数据集上进行消融实验并与现有实时语义分割网络进行对比,验证了各模块的有效性,以124.5 f/s达到了77.9%的平均交并比(mean intersection over union,MIoU);在CamVid数据集上以211.1 f/s达到了75.1%的MIoU。相比现有的实时语义分割网络,本文算法更好地权衡了分割的精度和速度。

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.

拉萨八廓历史街区文脉空间特征图谱研究

我国城镇化已处于换挡提质的下半程,城市地域文化特色消失、文脉断裂等问题也逐渐凸显。街区作为城市核心区域,识取街区特有文化符号有助于保护特色空间不在城镇化建设冲击下消失。本研究以拉萨城市核心区域——八廓历史街区为研究对象,通过引入城市形态学、建筑类型学分析方法,对街区“道路系统”和“建筑类型”两个层面的文脉空间特征进行“历时性”和“共时性”分析,构建街区文脉空间特征图谱,为其他类似地区的历史街区文脉空间可视化研究提供一定参考。

基于空间句法的城市更新空间形态研究——以湖州东街为例

城市更新是个动态的过程。湖州东街历史悠久,是反映古城依水而建的街巷肌理代表,既是历史城区中重要的视觉通廊,也是现今重要的商业街区。历史上的东街是商贸、行政、私家园林、官邸等重要建筑的分布地,而今的东街历史性逐渐消失,沿街业态混杂,空间结构松散,空间活力不足。文章利用空间句法原理,分析其空间形态,探究其空间发展趋势和特点,并且通过梳理空间结构,为城市更新改造献计献策,留住文脉,协调城市整体空间格局发展。

建筑类型学视角下重庆市地铁前区空间设计中城市文脉融入策略

随着物质生活水平不断提高,人们的出行方式发生较大变化,各城市相继开展地铁建设,但当前城市地铁前区空间面临同质化问题。故以重庆市地铁为例,从建筑类型学视角出发,分析问题产生原因,探索地铁前区空间设计中城市文脉融入策略,为重庆市地铁前区空间发展提供思路。

新时代“五育并举”下高校学习空间体系构建

2018年全国教育大会上,习近平总书记提出“德智体美劳全面发展”的教育目标。“五育并举”成为新时代高校教育改革的重要举措,从而为高校学习空间的规划和建设提出新命题。该文在已有研究基础上进一步阐述“五育并举”内涵,并结合正式-非正式学习空间分类,剖析实现“五育”的教育目标、教育活动及其对应的正式-非正式学习空间,试图构建“五育并举”新阶段下高校学习空间体系,从而为新时代中国特色社会主义高校校园空间规划与营造提供一定启示。

自动全局上下文感知相关滤波器跟踪算法

在相关滤波器跟踪算法中引入正则化后可以有效提高跟踪效率,但需要花费大量精力调整预定义参数,此外还有目标响应发生在非目标区域会导致跟踪漂移等问题,因此提出一种自动全局上下文感知相关滤波器(Automatic Global Context Awareness Correlation Filter,AGCACF)跟踪算法.首先,在跟踪过程中利用目标局部响应变化实现自动空间正则化,将自动空间正则化模块加入目标函数,使滤波器专注于目标对象的学习;其次,跟踪器利用目标全局上下文信息,结合自动空间正则化,使滤波器能及时学习到更多与目标有关的信息,减少背景对跟踪性能的影响;接着,在滤波器中加入时间正则化项,来充分学习目标在相邻帧之间的变化,从而获得更准确的模型样本.实验结果表明,与其他跟踪算法相比,AGCACF跟踪算法在距离精度和成功率方面具备更好的跟踪效果.

基于隐含空间知识融合的输电线路多金具检测方法

针对输电线路多金具检测任务中的小目标问题和密集遮挡问题,提出基于隐含空间知识融合的输电线路多金具检测方法。首先,为了挖掘输电线路金具间的隐含空间知识以协助模型进行检测,提出空间框设定模块和空间上下文提取模块进行空间框的设定以及空间上下文信息的提取。然后,设计空间上下文记忆模块对空间上下文信息进行筛选和记忆,并由此辅助多金具检测模型的定位。最后,改进模型后处理部分以进一步缓解金具密集遮挡带来的低检测精度问题。实验结果表明,该模型对多类金具的检测有提升效果,尤其对于小目标金具和密集遮挡金具的提升尤为显著。且相比于基线模型,在整体的AP^(50)评价指标和更严格的AP^(75)评价指标上分别提高了3.5%和5.7%。这为后续金具检测的落地应用和进一步的故障诊断奠定了基础。

基于触媒理论的城市集市空间微更新设计研究——以西仓集市为例

目的建立触媒理论视角下典型集市空间微更新方法,探索基于城市触媒理论的城市集市空间保护与更新的设计路径,优化集市空间设施利用策略。方法通过对城市集市空间研究进展的梳理分析,确立集市空间在微更新视角下介入触媒设计方法的可能性和积极影响;通过对触媒理论概念的阐述,以空间要素、功能要素和文化要素作为原始触媒因子,构建城市触媒理论介入集市空间微更新的设计方法;以西安西仓集市街区微更新设计为例,探索触媒载体与城市集市活力塑造的结合点,通过系列化和多样化的微小干预,激发街区发展的“链式反应”,探索实现集市空间环境的活化。结论传统集市街区介入触媒理论以“空间、功能、文化”三要素触媒因子植入集市空间的路径,通过空间形态重构、街区功能置换、历史脉络修复等有机结合,可以达到低冲击导向下的城区微更新的整体发展目标,提升环境活力、优化设施效率、引导空间活化,为集市街区改造再利用提供新的设计研究视角。

基于多尺度语义编解码网络的遥感图像语义分割

针对遥感图像语义分割中存在的多层次信息提取和多尺度特征图上下文依赖性两个问题,本文分析现有处理方案,提出了一种综合运用多项技术的多尺度语义编解码网络(Multi-scale Semantic Encoder-Decoder Networks,MSEDNet).MSEDNet由编码与解码两部分构成.编码阶段,首先提出残差协同空间注意(Residuals Coordinate Spatial Attention,RCSA)的MobileNetV3增强型模块,提取语义信息;其次,设计多层增强语义上下文模块(Enhance Semantic Context Module,ESCM),提升多尺度结构特征图的表征能力.解码阶段,首先提出多核卷积与Focus并行的强化空间细节信息模块(Strengthen Spatial Detail Information Module,SSDIM),增强浅层特征细节和结构信息;其次,设计了三元迭代多尺度特征融合(Triplet Iterative Multi-Scale Feature Fusion,TIMSFF)策略,强化图像深层全局语义信息与浅层局部细节特征的多尺度融合,提升分割精度.所提模型在ISPRS Vaihingen和Potsdam数据集上验证,总体分割精度(Overall Accuracy,OA)分别达到95.699%、95.534%,平均F1-score(mean F1-score,m F1)分别提高2.661%和2.929%,且平均交并比(mean Intersection over Union,mIoU)分别增长3.973%和4.012%.所耗参数量Param下降至6.77 M.

Granger因果关系时空图推理的群体行为分析

因果关系普遍存在于群体交互行为中,体现出主动体行为对被动体行为的有向影响.因果关系检测的难点在于交互双方的行为具有复杂的时间动态性.现有方法使用循环神经网络,来描述交互关系的时间变化特性,并使用时间注意力机制,来描述时间依赖关系.上述方法忽视了对多人依赖关系的分析,难以区分交互双方中的主动行为者和被动行为者.本文设计了一种基于Granger因果关系的时空图推理模型,来学习交互双方的主动和被动关系.为了实现Granger因果关系检测,该模型对单个个体时序特征进行自回归建模,来描述行为对个体自己的依赖.该模型对两个个体时序特征进行相关回归建模,来描述行为对两个个体的依赖.该模型通过比较自回归误差和相关回归误差,当自回归误差明显大于相关回归误差,则说明相关个体改变了另一方个体的行为特征,从而检测出相关个体为主动个体,另一方为被动个体.相关回归模型考虑了多种时间延迟量的两个个体的时序特征序列,用于学习两个个体之间行为的时间延迟量.该时间延迟量用于将主动个体时间特征与被动个体时间特征进行对齐.时间对齐后的主动个体特征提供了被动个体的时间和空间上下文特征,并与被动个体特征进行通道级的融合.为了充分描述个体之间的外观模式,位置约束,因果关系的交互关系,该模型构建多尺度外观的因果图,并使用图推理学习融合上下文的个体特征和群体特征.本文对Granger因果关系检测进行消融分析,并说明时间延迟量,交互融合通道比例,多尺度图推理,能够有效改善个体特征、群体特征的描述能力.本文方法在Volleyball和Collective Activity数据集上优于现有群体行为识别方法.本文的可视化结果说明Granger因果关系可以捕获群体中关键的交互关系.