User‐Centric Conflict Management for Media Services Using Personal Companions

In this paper, we propose a user‐centric conflict management method for media services which exploits personal companions for the harmonious detection and resolution of service conflicts. To detect conflicts based on the varying characteristics of individual users, the proposed method exploits the unified context describing all users attempting to access media services. It recommends and mediates users' preferred media contents through a shared screen and personal companions to resolve the detected conflicts. During the recommendation, a list of preferred media contents is displayed on the shared screen, and a personally preferred content list is shown on the user's personal companion comprising the selection of media contents. Mediation assists the selection of a consensual service by gathering the users' selections and highlighting the common media contents. In experiments carried out in a ubiHome, we observed that recommendations and mediation are useful in harmoniously resolving conflicts by encouraging user participation in conflict situations.     

基于环境状态的移动计算

首先简要介绍了移动计算的发展和面临的问题 ,提出了基于环境状态的移动计算 ;然后解释了什么是环境状态以及其分类 ,并从环境状态出发 ,阐述了基于环境状态的移动计算的特点和类型 ;其中 ,给出了目前一些基于环境状态的移动计算的应用实例 ,以位置环境状态为例 ,提出了基于环境状态的移动计算的基本模块和所完成的功能 ;最后 ,从两种计算模型出发 ,说明目前应用的状况及以后基于环境状态的移动计算的发展     

引入全局上下文特征模块的DenseNet孪生网络目标跟踪

近年来,采用孪生网络提取深度特征的方法由于其较好的跟踪精度和速度,成为目标跟踪领域的研究热点之一,但传统的孪生网络并未提取目标较深层特征来保持泛化性能,并且大多数孪生网络只提取局部领域特征,这使得模型对于外观变化是非鲁棒和局部的。针对此,该文提出一种引入全局上下文特征模块的DenseNet孪生网络目标跟踪算法。该文创新性地将DenseNet网络作为孪生网络骨干,采用一种新的密集型特征重用连接网络设计方案,在构建更深层网络的同时减少了层之间的参数量,提高了算法的性能,此外,为应对目标跟踪过程中的外观变化,该文将全局上下文特征模块(GC-Model)嵌入孪生网络分支,提升算法跟踪精度。在VOT2017和OTB50数据集上的实验结果表明,与当前较为主流的算法相比,该文算法在跟踪精度和鲁棒性上有明显优势,在尺度变化、低分辨率、遮挡等情况下具有良好的跟踪效果,且达到实时跟踪要求。

     

基于局部相似的红外小目标跟踪算法

连续图片序列中的运动对象与其局部背景保持相对稳定.相邻图像序列的目标局部灰度分布具有相似性.针对红外小目标跟踪杂波干扰与运动模糊问题,设计了基于局部相似的目标增强方法,并提出一种基于局部相似和运动估计的跟踪算法.该算法根据目标的局部相似性构建搜索空间,采用运动估计机制缩小搜索域,然后利用时空上下文学习跟踪算法产生跟踪结...     

概念格在二值命题逻辑命题集约简中的应用

概念格的属性约简理论和命题逻辑系统中命题集的约简理论似乎是独立发展的两个研究分支,本文在二值命题逻辑中引入由命题集Γ所诱导的形式背景的概念,并基于此建立了概念格;在二值命题逻辑中提出了命题集Γ的约简理论,即在保持Γ推理能力不变的前提下对Γ中的成员进行约简;运用概念格的方法从Γ及其子集的关系出发给出了Γ约简的判定定理以及求Γ约简的方法.     

SIFT与形状上下文结合的异源图像匹配算法

针对红外图像和可见光图像灰度差异大、匹配困难的问题,提出了一种基于改进的尺度不变特征变换和形状上下文描述的局部多特征匹配算法。首先通过高斯差分检测算法分别提取两幅图像的特征点;针对特征点梯度方向存在反转现象,结合梯度镜像方法对特征点统计特征点邻域梯度方向信息;然后引入图像边缘特征生成形状上下文描述子,与梯度方向描述子级联成联合描述子;最后采用欧氏距离和卡方距离加权的联合距离和最近邻算法对特征点进行匹配。实验结果证明,在红外图像和可见光图像匹配中,该算法相比原始SIFT算法能有效减少误匹配特征点对,达到较高的匹配精度。     

本体映射的集成和语义挖掘研究

对于不同的待映射本体,各种映射策略的表现可能有较大差异,传统的映射方法并没有根据具体语境和策略的可信度进行映射结果集成,对产生的映射对也没有做进一步的语义挖掘。为了解决这个问题,根据本体和谐度对各策略权值进行自适应计算;利用结构语义信息进行语义传播(或利用在线本体作为背景知识)挖掘新的映射对。实验表明改进的方法可以明显提高查准率,获得更准确的映射结果。     

Does collaborative and experiential work influence the solution of real-context estimation problems? A study with prospective teachers

Although it is a challenge for primary school teachers, real-context estimation problems can be used as an introduction to mathematical modeling. With this aim, we designed a two-phase activity: in the first phase, 224 prospective teachers developed individual action plans to solve a sequence of real-context estimation problems in the classroom; in the second phase, they completed the solution of the same problems working in groups in the real location where the four problems were contextualized. A comparative study showed that, in the second phase, prospective teachers were able to adapt their solutions to contextual features detected in situ that had not been anticipated in the action plans developed during the first phase. Two-phase modeling activities, which permit a comparison of different perspectives on problems, facilitate the experience of collaborative work. These activities could be incorporated into prospective teachers’ initial training as a useful resource for improving their problem-solving expertise.     

基于TLD框架的上下文目标跟踪算法

提出了一种基于TLD (Tracking-Learning-Detection)框架的上下文目标跟踪算法.在TLD框架中,融入时空上下文跟踪算法,提高跟踪器的鲁棒性和稳定性.引入Kalman滤波来处理目标被严重遮挡时跟踪失效的问题.此外,采用由粗到精的搜索策略进行目标检测,利用帧差法确定运动目标疑似区域,提高检测效率.实验结果表明所提出的算法具有较好的鲁棒性和实时性.     

Context-aware end-to-end QoS diagnosis and quantitative guarantee based on Bayesian network

To support quality of service(QoS)management on current Internet working with best effort, we bring forth a systematic approach for end-to-end QoS diagnosis and quantitative guarantee. For QoS diagnosis, we take contexts of a service into consideration in a comprehensive way that is realized by exploiting causal relationships between a QoS metric and its contexts with the help of Bayesian network(BN)structure learning. Context discretization algorithm and node ordering algorithm are proposed to facilitate BN structure learning. The QoS metric is diagnosed to be causally related to its causal contexts, and the QoS metric can be quantitatively guaranteed by its causal contexts. For quantitative QoS guarantee, those causal relationships are first modeled quantitatively by BN parameter learning. Then, the QoS metric is guaranteed to certain value with a probability given its causal contexts tuned to suitable values, that is, quantitative QoS guarantee is reached. Simulations with three sequential stages:context discretization, QoS diagnosis and quantitative QoS guarantee, on a peer-to-peer(P2P)network, are discussed and our approach is validated to be effective.