基于可编程图形硬件的实时阴影算法及实现

阴影在增强三维场景真实感方面起着非常重要的作用。随着计算机图形硬件的发展,利用可编程图形硬件加速并提高渲染效果已经成为主要发展趋势。基于当前可编程图形硬件技术,针对目前shadow maps算法存在的问题,提出了一种根据场景特征程序自动生成合适偏移量的实时的阴影生成算法,该算法很好的改善了阴影渲染效果,防止了走样。最后给出了算法基于GLSL着色语言的实现代码。     

融合强化学习和关系网络的样本分类

针对少量训练样本在深度学习算法中难以实现高精度分类的问题,提出一种融合强化学习和关系网络的小样本分类算法。采用图像预处理过程中基于强化学习的美学意识图像自动裁剪模型,通过构建美学意识奖励函数来输出最佳裁剪图像,从而保留图像最具特征部分。利用关系网络模型,将自动裁剪后的小样本图像中的训练样本图像与测试图像通过关系网络中的嵌入模块进行特征提取。将提取后的特征进行特征映射级联,并将级联后的特征映射馈送到关系网络中的关系模块中进行比较,将最终产生的0到1范围内的关系评分作为比较结果,从而判断测试图像所属的类别。在小样本数据集上进行实验并与现有方法进行对比,实验表明该方法能够实现较高精度的小样本分类。     

Building ontology based knowledge maps to assist business process re-engineering

Business Process Re-engineering (BPR) is being used to improve the efficiency of the organizational processes, however, a number of obstacles have prevented its full potential from being realised. One of these obstacles is caused by an emphasis on the business process itself at the exclusion of considering other important knowledge of the organization. Another is due to the lack of tools for identifying the cause of the inefficiencies and inconsistencies in BPR. In this paper we propose a methodology for BPR that overcomes these two obstacles through the use of a formal organizational ontology and knowledge structure and source maps. These knowledge maps are represented formally to facilitate an inferencing mechanism which helps to automatically identify the causes of the inefficiencies and inconsistencies. We demonstrate the applicability of this methodology through the use of a case study of a university domain.     

基于密度信息的改进降维方法

扩散映射(diffusion maps)是一种基于流形学习的非线性降维方法。为了提高降维的效果, 根据近邻点的选取对diffusion maps的降维效果影响, 利用数据近邻点分布的不同, 挖掘该数据点局部的密度信息, 能够更好地保持数据的流形结构。利用样本点聚类后的类别信息构造密度信息指数, 提出了一种改进的diffusion maps算法, 有效地保持了高维数据中的流形结构, 所提的新算法在多种实验中得到了证实。     

基于混沌消息扩展的Hash函数

构造了一种基于混沌消息扩展的Hash函数,方案中将消息权重矩阵融入时空混沌系统,加强了每个消息块中各字符之间及各消息块之间的联系,构建了统计特性良好的扩展消息,采用4轮64步迭代得到160 bit的消息摘要。和SHA-1算法相比在不明显影响运算效率的前提下安全性得到了提高。     

Learning multiple experiences useful visual features for active maps localization in crowded environments

Crowded urban environments are composed of different types of dynamic and static elements. Learning and classification of features is a major task in solving the localization problem in such environments. This work presents a gradual learning methodology to learn the useful features using multiple experiences. The usefulness of an observed element is evaluated by a scoring mechanism which uses two scores – reliability and distinctiveness. The visual features thus learned are used to partition the visual map into smaller regions. The robot is efficiently localized in such a partitioned environment using two-level localization. The concept of active map (AM) is proposed here, which is a map that represents one partition of the environment in which there is a high probability of the robot existing. High-level localization is used to track the mode of the AMs using discrete Bayes filter. Low-level localization uses a bag-of-words model to retrieve images and accurately localize the robot. The pose of the robot is the one retrieved from the AM that has maximum a posteriori. Experiments have been conducted on a unique highly crowded data-set collected from Indian roads. The results support the proposed method due to speed and localization accuracy.     

Uniform Approximation of Discrete-Time Periodically Varying Systems

It is shown that the elements of a large class of discrete-time periodically varying nonlinear input-output maps can be uniformly approximated arbitrarily well, over infinite time intervals, using a certain structure that can be implemented in many ways using, for example, radial basis functions, polynomial functions, piecewise linear functions, sigmoids, or combinations of these functions. For the special case in which these functions are taken to be certain polynomial functions, the input-output map of our structure is a generalized finite discrete-time Volterra series. Results are given for the case in which inputs and outputs are defined on the integers. The case in which inputs and outputs are defined on the nonnegative integers is also addressed.     

一种数字影像地图无损水印方案(英文)

A lossless digital watermarking scheme,which was zero- perturbation on the content and graphics for the digital image maps,is proposed.During the simulation,the scheme has been utilized to model the copyright protection program as a commerce application.Compared to the traditional digital watermarking schemes,our scheme operates in redundancy areas of maps and is scalable to topology changes.Experimental results show that,with respect to the geometric attacks and image transformed,the performance of our scheme is better than the classical algorithms based space or frequency domain with much lower complexity..     

卡诺图在数字电路分析和设计中的应用

卡诺图是数字电路中非常重要的分析和设计工具。通过若干实例,揭示了卡诺图的直观性和实用性,系统全面地掌握卡诺图的各种应用,可以大大简化数字电路分析和设计的过程,起到事半功倍的效果。     

Vibrations of an aramid anchor cable subjected to turbulent wind

The vibration response of an initially pre-stressed anchor cable made of parallel-lay aramid fibres excited by a measured and artificially simulated spatial turbulent wind field is presented in the paper. Results of the analyses of in situ measured wind records are described. For selected data set statistical characteristics and power spectral density functions of the measured wind velocity components are calculated. The wind stochastic velocity fluctuation is modelled as a one-variate bi-dimensional random field. Cross-power spectral density functions, at different point locations are introduced. The combination of the weighted amplitude wave superposition method (WAWS) with the Shinozuka–Deodatis method is used for the analyzed problem. A time-dependent behaviour of the synthetic cable is investigated which is subjected to turbulent wind with large expected oscillations that arise as a result of slackening due to the relaxation effects. A nonlinear transient dynamic analysis is used in conjunction with the finite element method to determine the dynamic response of the cable subjected to turbulent wind at its initially prestressed state and in the selected times after the relaxation effect. The constitutive equation of the relaxation of the aramid cable follows an experimentally obtained law of the logarithmic type. To monitor the dependences of the individual quantities of cable vibration in the phase space, attractors and Poincaré maps are created by sampling the cable’s displacement and velocity at periods of relevant frequencies. Interesting findings based on the response of the cable with rheological properties to turbulent wind are presented.