A scalable, locality aware, fault-tolerant, and decentralized location scheme for distributed networks

This paper presents Isotope, an efficient, locality aware, fault-tolerant, and decentralized scheme for data location in distributed networks. This scheme is designed based on the mathematical model of decentralized location services and thus has provable correctness and performance. In Isotope, each node needs to only maintain linkage information with about O(log n) other nodes and any node can be reached within O(log n) routing hops. Compared with other related schemes, Isotope’s average locating path length is only half that of Chord, and its locating performance and locality-awareness are similar to that of Pastry and Tapestry. In addition, Isotope is more suitable for constantly changing networks because it needs to exchange only O(log n) O(log n) messages to update the routing information for nodes arrival, departure and failure.

AD Hoc网络中的一种QoS-AWARE多径路由协议

随着多媒体应用的不断发展,在Ad Hoc网络中提供QoS保证具有重要的意义。在研究节点拥塞、节点碰撞等本地信息对QoS路由影响的基础上,提出了一种在Ad Hoc网络中提供QoS保证的路由协议——基于本地信息的QoS-aware多径路由协议QAMR。该协议把带宽作为QoS参数,通过引入节点利用因子和路径利用因子来考虑可行路径处理能力问题和对业务的QoS保证。该协议的路由发现和路由维护算法是DSR协议的扩展。通过仿真和对比分析验证了QAMR协议无论在静态网络和动态网络,都具有很好的性能。

Awareness of and attitudes towards infertility and its treatment： a cross-sectional survey of men in a United States primary care population

Previous studies have described racial and socioeconomic disparities in the treatment of infertility. Patient factors such as attitudes and awareness may be contributing factors. Since primary care is often the setting that serves as an entry into other areas of medicine, we sought to evaluate men＇s attitudes and awareness of male infertility in the primary care setting. To do this, we performed a cross-sectional survey of men＇s attitudes toward men＇s health issues in 210 men from two primary care clinic waiting rooms in Atlanta, Georgia. The survey was self-administered with closed-ended question items and was approximately 20 min in length. Of the 310 men approached, 210 agreed to participate and returned completed surveys. Overall, 52% of men said they were ＂very＂ or ＂somewhat＂ familiar with infertility and 25% were familiar with treatments for infertility. Some men had heard of surgery （21%） and medication （35%） as treatments for male infertility. Awareness and familiarity with the condition was greater in high socioeconomic status men （i.e. college graduates or those with income 〉$100 k per year） but did not differ by race on multivariate analysis. Attitudes toward infertility varied by race with non-Caucasian men being more likely to indicate that infertility is a serious condition, to be concerned about infertility, and to believe it decreases a man＇s quality-of-life. Therefore, a lack of awareness, but not negative attitudes, may contribute to previously-described disparities in the treatment of infertility.

An Efficient Detection Approach of Content Aware Image Resizing

Content aware image resizing(CAIR)is an excellent technology used widely for image retarget.It can also be used to tamper with images and bring the trust crisis of image content to the public.Once an image is processed by CAIR,the correlation of local neighborhood pixels will be destructive.Although local binary patterns(LBP)can effectively describe the local texture,it however cannot describe the magnitude information of local neighborhood pixels and is also vulnerable to noise.Therefore,to deal with the detection of CAIR,a novel forensic method based on improved local ternary patterns(ILTP)feature and gradient energy feature(GEF)is proposed in this paper.Firstly,the adaptive threshold of the original local ternary patterns(LTP)operator is improved,and the ILTP operator is used to describe the change of correlation among local neighborhood pixels caused by CAIR.Secondly,the histogram features of ILTP and the gradient energy features are extracted from the candidate image for CAIR forgery detection.Then,the ILTP features and the gradient energy features are concatenated into the combined features,and the combined features are used to train classifier.Finally support vector machine(SVM)is exploited as a classifier to be trained and tested by the above features in order to distinguish whether an image is subjected to CAIR or not.The candidate images are extracted from uncompressed color image database(UCID),then the training and testing sets are created.The experimental results with many test images show that the proposed method can detect CAIR tampering effectively,and that its performance is improved compared with other methods.It can achieve a better performance than the state-of-the-art approaches.

Burstiness-Aware Congestion Control Protocol for Wireless Sensor Networks

In monitoring Wireless Sensor Networks(WSNs),the traffic usually has bursty characteristics when an event occurs.Transient congestion would increase delay and packet loss rate severely,which greatly reduces network performance.To solve this problem,we propose a Burstiness-aware Congestion Control Protocol(BCCP) for wireless sensor networks.In BCCP,the backoff delay is adopted as a congestion indication.Normally,sensor nodes work on contention-based MAC protocol(such as CSMA/CA).However,when congestion occurs,localized TDMA instead of CSMA/CA is embedded into the nodes around the congestion area.Thus,the congestion nodes only deliver their data during their assigned slots to alleviate the contention-caused congestion.Finally,we implement BCCP in our sensor network testbed.The experiment results show that BCCP could detect area congestion in time,and improve the network performance significantly in terms of delay and packet loss rate.

The Knowledge about the Potential Health Risks of Illegal Bushmeat Activities among Local Communities Adjacent to Western Nyerere National Park, Tanzania

Establishing knowledge of local communities regarding the potential health risks of illegal bushmeat hunting and its related practices is among the essential means to developing effective conservation and public health programs. To reveal the understanding of the local people regarding the potential health risks of bushmeat consumption and handling of wildlife products, we used a semi-structured questionnaire to survey 261 households in eight villages located adjacent to the western part of Nyerere National Park. Also, we interviewed eight key informants, including conservation personnel and veterinary officers, in the study area. The proportion of local people who were unaware that handling of wildlife products and bushmeat consumption were risky behaviours towards acquiring zoonotic diseases was slightly higher (57%) than the proportion of respondents who were aware of the likely risks of zoonotic diseases from the practices. After all, the majority (83%) of local people admitted having come into contact with wildlife products, while over 70% reported having consumed bushmeat. We found that local communities living closer to the park boundaries (<15 km) have a higher likelihood of contacting wildlife products, and that community members who have lived in the area for more than ten years have experienced more contact with wildlife products than immigrants. Moreover, the age of inhabitants and length of stay were found to be the most significant factors in determining the likelihood of bushmeat consumption. Most people seemed to prefer bushmeat to domestic meat, arguing that bushmeat is far more flavorful, tender, not tainted with chemicals, and has less fat, making it safer. Our study demonstrated that local communities participate in risky practices without contemplating the health consequences that could emerge should a zoonotic disease outbreak occur. Among other strategies, we recommended conservation and public health institutions to increase awareness campaigns on the possible health risks of zoonotic diseases associated with wildlife.

区域感知校准的自适应人群计数与定位方法

密集场景下准确人群计数和定位,对于保障公共安全具有重要的意义。针对密集人群计数与定位易受人群分布不均、背景干扰等因素的影响,导致计数定位不准确的问题,提出一种基于区域感知校准的自适应人群计数与定位方法。通过构建金字塔结构提取人群图像的多尺度特征,增强特征关联性,并设计可变形几何自适应模块学习不同分布的人群几何特征,以增强对人群分布不均的适应性。在此基础上,提出区域感知和区域校准模块,提取全局上下文特征和区域特征,克服了背景干扰造成的定位与计数不准问题。接着通过双分支卷积预测通路,输出生成点的预测位置和置信度分数,以提高网络的定位与计数精度。最后提出改进二分图最大匹配Hopcroft-Karp算法对真值点与预测点进行匹配校准,从而完成人群定位与计数。实验结果表明,所提方法分别在公开的ShanghaiTech Part A和Part B数据集、NWPU-Crowd数据集、UCF-QNRF数据集上评价指标均优于对比算法,且定位精度较P2Pnet分别提高了3.5%、6.1%、11.3%和8.1%,能够有效提高人群定位与计数的准确度。

比例融合与多层规模感知的人群计数方法

针对密集场景下人群图像拍摄视角或距离多变造成的多尺度特征获取不足、融合不佳和全局特征利用不充分等问题,提出一种比例融合与多层规模感知的人群计数网络。首先采用骨干网络VGG16提取人群密度初始特征;其次,设计多层规模感知模块,获得人群多尺度信息的丰富表达;再次,提出比例融合策略,根据卷积层捕获的特征权重重构多尺度信息,提取显著性人群特征;最后,采用卷积回归策略进行密度图的回归。同时,提出一种局部一致性损失函数,通过区域化密度图的方式增强生成密度图与真实密度图的相似度,提高计数性能。在多个人群数据集上的试验结果表明,所提模型优于近年人群计数的先进方法,且在车辆计数上有较好推广性。

整体与局部相互感知的图网络时序动作检测

针对目前动作检测与定位方法未综合利用整体与局部相互感知的时空关系信息、不利于提升动作检测与定位性能的问题,提出整体与局部相互感知的图网络时序动作检测方法.该方法综合利用各动作提案的特征相似性和时序重叠度构建整体关系图推理子网络,通过学习获得提案,该提案包含更丰富的整体时空特征表示;利用提案发生的时间偏序关系,构建局部关系图推理子网络,该子网络包含多个级别三体相似图和三体互补图的结构,通过学习获得不同时间尺度下提案的局部关系信息;最后构成整体与局部关系相互感知的丰富特征表达,用于动作检测与定位.采用平均精度均值作为评价指标在2个公开数据集(Thumos14和ActivityNet1.3)上进行了试验.结果表明,与PGCN、G-TAD、TAL-Net、CDC等先进方法相比,文中方法能有效提高动作检测的性能.

民族服装图像描述生成的局部属性注意网络

针对民族服装图像属性信息复杂、类间相似度高且语义属性与视觉信息关联性低,导致图像描述生成结果不准确的问题,提出民族服装图像描述生成的局部属性注意网络.首先构建包含55个类别、30000幅图像,约3600 MB的民族服装图像描述生成数据集;然后定义民族服装208种局部关键属性词汇和30089条文本信息,通过局部属性学习模块进行视觉特征提取和文本信息嵌入,并采用多实例学习得到局部属性;最后基于双层长短期记忆网络定义包含语义、视觉、门控注意力的注意力感知模块,将局部属性、基于属性的视觉特征和文本编码信息进行融合,优化得到民族服装图像描述生成结果.在构建的民族服装描述生成数据集上的实验结果表明,所提出的网络能够生成包含民族类别、服装风格等关键属性的图像描述,较已有方法在精确性指标BLEU和语义丰富程度指标CIDEr上分别提升1.4%和2.2%.

基于GAN的肺部CT影像超分辨率重建研究

基于提高CT图像分辨率、丰富放大后图像纹理细节的目的,提出了RUAGAN模型。通过在生成器部分使用RRDB作为基本块,并加入局部注意力(Local aware Attention,LA)的方法提取影像中的高频信息;为提供更加真实的梯度反馈,使用U-Net网络作为鉴别器。经过实验仿真表明,改进后的模型在肺部CT图像数据集上训练,在4x放大因子上重建出的图像纹理更加丰富,有清晰真实的边缘,其峰值信噪比与结构相似性分别达到31.980和0.974,高于其他分辨率重建模型。

地方应用型高校转型过程中教师发展的困境与突破

教师是教育教学活动的中坚力量,地方应用型高校对教师的教育观念更新、知识框架重构和产教融合能力提升等提出了新的要求。目前,地方应用型高校教师发展存在主体定位模糊下的专业认知转型、认知转型困惑下的专业知识重构和社会服务经历缺乏下的实践技能生成等困境。推进地方应用型高校教师发展,促进教师专业知识革新与实践技能重构,亟需觉醒教师转型意识,加强技能实践落地,突破产学研共融壁垒。

Geographical Accessibility to Basic Infrastructures and Services in Rural Senegal: The Case of the Niakhar Area

Access to basic infrastructure and services is a factor in economic development and an important aspect in combatting social and spatial disparities. But this access is often subject to several constraints, including geographical accessibility. In this article, we aim to analyze the geographical accessibility to basic infrastructure and services in the Niakhar area, using the improved two step floating catchment area method and local spatial association indicators. The results reveal that the areas with high accessibility to health and education infrastructures and services are mainly located along the south-east and north-west gradient, while those with low accessibility are found in the south-west and north-east center. They also show high accessibility to trade services in the center of the study area.

面向跨视角地理定位的感知特征融合网络

跨视角地理定位是指同一地理目标可通过检索多个平台视角(无人机、卫星和街景)进行位置定位。这类定位任务主要挑战是不同视角点间的剧烈变化,降低了模型的检索性能。目前跨视角地理定位的网络存在以下的问题。由于地理目标具有尺度和角度的多样性,当前网络在感知目标信息时容易受到局部区域的干扰。属于同一类别中的不同视角,它们的角度差异很大。因此,提出了面向跨视角地理定位的感知特征融合网络(PFFNet)来学习位置感知特征并在每个视角之间建立语义关联。在PFFNet中的每个视角,搭建分流上下文嵌入网络(SCENet)作为骨干网络分别提取每个视角的上下文关联特征信息并构建目标位置的编码空间。在跨视角地理定位数据集University-1652上,将提出的方法与最先进的方法进行比较。实验结果表明,所提出感知特征融合网络在大规模数据集中取得了较高的自适应性能。

基于高分辨率类激活映射算法的弱监督目标实时检测

受益于深度学习的发展,目标检测技术在各类视觉任务中得到广泛关注。然而,获取目标的边框标注需要高昂的时间和人工成本,阻碍了目标检测技术在实际场景中的应用。为此,该文在仅使用图像类别标签的基础上,提出一种基于高分辨率类激活映射算法的弱监督目标实时检测方法,降低网络对目标实例标注的依赖。该方法将目标检测细划分为弱监督目标定位和目标实时检测两个子任务。在弱监督定位任务中,该文利用对比层级相关性传播理论设计了一种新颖的高分辨率类激活映射算法(HR-CAM),用于获取高质量目标类激活图,生成目标伪检测标注框。在实时检测任务中,该文选取单镜头多盒检测器(SSD)作为目标检测网络,并基于类激活图设计目标感知损失函数(OA-Loss),与目标伪检测标注框共同监督SSD网络的训练过程,提高网络对目标的检测性能。实验结果表明,该文方法在CUB200和TJAB52数据集上实现了对目标准确高效的检测,验证了该文方法的有效性和优越性。

基于数据挖掘的企业局域网安全态势感知方法

为提高企业局域网运行的安全性与稳定性,准确感知和应对潜在的网络威胁,开展基于数据挖掘的企业局域网安全态势感知方法研究。首先,采集并预处理企业局域网安全数据。然后,利用数据挖掘算法,设计关联规则挖掘,从安全数据中提取网络态势特征。最后,构建态势评估模型,计算企业局域网安全态势的基本指数,评估当前企业局域网的态势情况,分析可能的威胁来源,确定威胁级别及对应威胁程度,实现安全态势感知。实验结果表明,该方法应用后,感知误报率均低于0.5%,能够准确识别出潜在的安全威胁,从而提高企业的网络安全防护能力。

西藏交通警察道路执法工作的风险与防范

西藏自治区自然、地理与气候环境复杂且特殊,道路交通环境相对恶劣,这导致当地交通警察道路执法工作面临诸多风险,包括警员人身安全和执法效能等风险。对此,文章对西藏交通警察道路执法工作风险进行了调查分析,从地区环境、警员安全意识与技能、执法装备、社会宣传和教育等角度探讨了相关风险的成因,针对性相关风险成因,提出从优化警力配置与道路信息推送、加强警员培训、完善警务装备、加强法治化宣传教育等工作来消除相关风险,以期保障警员安全并改善西藏交通环境。

Local algorithms on dense two-frame stereo matching

Stereo matching is a fundamental and crucial problem in computer vision. In the last decades, many researchers have been working on it and made great progress. Generally stereo algorithms can be classified into local methods and global methods. In this paper, the challenges of stereo matching are first introduced, and then we focus on local approaches which have simpler structures and higher efficiency than global ones. Local algorithms generally perform four steps： cost computation, cost aggregation, disparity computation and disparity refinement. Every step is deeply investigated, and most work focuses on cost aggregation. We studied most of the classical local methods and divide them into several classes. The classification well illustrates the development history of local stereo correspondence and shows the essence of local matching along with its important and difficult points. At the end we give the future development trend of local methods.

Motion Localization with Optic Flow for Autonomous Robot Teams and Swarms

The ability to localize moving objects within the environment is critical for autonomous robotic systems. This paper describes a moving object detection and localization system using multiple robots equipped with inexpensive optic flow sensors. We demonstrate an architecture capable of detecting motion along a plane by collecting three sets of one-dimensional optic flow data. The detected object is then localized with respect to each of the robots in the system.

How can home care realize people-based, local and global consciousness of the organic unity of vision

Under the background of China＇s aging population, home care service system in line with China＇s traditional family pension, pension also incorporates community features, is widely used in China at this stage of pension model. However, the home care service research stage and the organic unity of the people did not do standard, local awareness and global perspective, often blindly follow the tide of globalization. Article by home care services, people-based, local and global vision of the concept of consciousness and its relationship to reports, summarized the research stage of home care service should be how to achieve unity of the three.