Eigensolution of specially structured matrices with hyper‐symmetry

Recently, four canonical forms have been developed and applied to the dynamics and stability analysis of symmetric frames. In this paper, hyper‐symmetric matrices and specially structured matrices are defined and efficient methods are proposed for the eigensolution of such matrices. Applications are extended to hyper‐graphs and specially structured graphs. Simple methods are developed for calculating the eigenvalues of the Laplacian matrices of such graphs. The developments presented in this paper can also be considered as generalization of Form II and Form III symmetry, previously defined by the authors. Copyright © 2006 John Wiley & Sons, Ltd.     

Laplacian group sparse modeling of human actions

Recently, many local-feature based methods have been proposed for feature learning to obtain a better high-level representation of human behavior. Most of the previous research ignores the structural information existing among local features in the same video sequences, while it is an important clue to distinguish ambiguous actions. To address this issue, we propose a Laplacian group sparse coding for human behavior representation. Unlike traditional methods such as sparse coding, our approach prefers to encode a group of relevant features simultaneously and meanwhile allow as less atoms as possible to participate in the approximation so that video-level sparsity is guaranteed. By incorporating Laplacian regularization the method is capable to ensure the similar approximation of closely related local features and the structural information is successfully preserved. Thus, a compact but discriminative human behavior representation is achieved. Besides, the objective of our model is solved with a closed-form solution, which reduces the computational cost significantly. Promising results on several popular benchmark datasets prove the efficiency and effectiveness of our approach.     

Sentiment visualization and classification via semi-supervised nonlinear dimensionality reduction

Sentiment analysis, which detects the subjectivity or polarity of documents, is one of the fundamental tasks in text data analytics. Recently, the number of documents available online and offline is increasing dramatically, and preprocessed text data have more features. This development makes analysis more complex to be analyzed effectively. This paper proposes a novel semi-supervised Laplacian eigenmap (SS-LE). The SS-LE removes redundant features effectively by decreasing detection errors of sentiments. Moreover, it enables visualization of documents in perceptible low dimensional embedded space to provide a useful tool for text analytics. The proposed method is evaluated using multi-domain review data set in sentiment visualization and classification by comparing other dimensionality reduction methods. SS-LE provides a better similarity measure in the visualization result by separating positive and negative documents properly. Sentiment classification models trained over reduced data by SS-LE show higher accuracy. Overall, experimental results suggest that SS-LE has the potential to be used to visualize documents for the ease of analysis and to train a predictive model in sentiment analysis. SS-LE can also be applied to any other partially annotated text data sets.     

Supporting the design process with hypergraph genetic operators

In this paper an evolutionary technique is proposed as a method for generating new design solutions with genotypes represented in the form of hierarchical hypergraphs. Such hypergraph-based evolutionary design requires the adaptation of traditional evolutionary operators like cross-over and mutation. This paper presents an attempt at defining modified cross-over and mutation operators that act on hierarchical hypergraphs. The application of the proposed transformations is illustrated on examples of designing floor layouts and furniture with the use of the evolutionary method.     

基于拉普拉斯方向的差值线性判别分析

标准的LDA方法通常有3个问题:1)为了确保类内散度矩阵的非奇异性,必须首先通过PCA进行维数约简,这限制了对更多维数空间的使用;2)当每人只有单个训练样本时,类内散度矩阵必然奇异,此时LDA无法工作;3)缺乏对像素间的局部相关性的考虑。为了解决这些问题,提出一种基于拉普拉斯方向的差值线性判别分析方法。该方法通过拉普拉斯方向实现更鲁棒的图像相异性测度,通过引入差值散度矩阵来避免类内散度矩阵的奇异性。实验结果显示,该算法对表情变化、光照改变及不同遮挡情况获得了更高的识别率,尤其针对光照变化,效果更加显著。     

PE-匹配图的可分性

为了得出一类密度矩阵的可分判据研究了特殊图,利用图理论、拉普拉斯矩阵性质、部分转置正判据、图上顶点与其部分转置图上对应顶点之间的度数关系,分别给出了完全纠缠(perfect entangled,PE)匹配图在C~pC~q与C~3C~4量子系统中的可分判据.证明了在C~pC~q量子系统中,若n=pq个顶点上的PE-匹配图的部分转置不是PE-匹配的,则该图的密度矩阵是纠缠的,否则其部分转置是非负(positive partialtranspose,PPT)的;并给出了C~3C~4系统中n=3×4个顶点上的PE-匹配图的密度矩阵可分的充要条件是该图的部分转置也是PE-匹配图.     

基于特征保持的三维模型四面体化及优化

三维模型四面体化是一种重要的有限元网格生成技术.介绍了一种特征保持的四面体网格生成及优化算法.首先使用三维模型主成分分析进行预处理,然后用体心立方构建初始四面体,接着通过拉普拉斯坐标改变模型边界切点的移动方式保持模型的局部特征,最后构造改进的密度能量误差函数优化四面体网格质量.实验结果表明,该方法可行、有效,且能很好地保持模型特征.     

基于量子力学和拉普拉斯金字塔的图像融合方法

针对数字图像的融合问题,研究一种基于量子力学理论的图像分解方法,并将其结合拉普拉斯金字塔变换,提出一种新的图像融合方法.先把2幅灰度图像表示成量子比特的形式,然后把每幅图像分解成4幅特征子图,根据分解后特征子图的不同含义,使用不同的融合策略分别对子图进行融合.其中一幅子图使用拉普拉斯金字塔变换进行融合,另外3幅子图使用区域梯度取大法进行融合,最后通过重构得到融合图像.仿真实验表明该方法融合效果较好,在主观视觉感受和客观评价指标上均优于传统的图像融合方法.     

带 2细分结构的四边形化

在Morse函数理论的基础上,提出一种新的从三角网格中建立四边形网格多分辨率表示的方法．先由人工指定或从拉普拉斯矩阵的特征函数中提取临界点,计算带约束的拉普拉斯方程得到光滑的Morse函数．函数的临界点（极大、极小和鞍点）有规律地分布在模型表面,在三角网格表面梯度场的引导下, 生成临界点间流线,得到临界点间的拓扑关系．通过临界点交换规则,同样是采用流线的方法,得到更精细的四边形网格。 最终可实现无需参数化而仅用流线方法来建立不同多分辨率表示的四边形网格。