结合颜色不变量的SIFT和形状上下文图像匹配算法

针对传统图像匹配算法sift和shape-context存在的不足,把这两种算法分别作了改进,并提出一种二者相结合的混合匹配算法。首先在传统sift算法的基础上融入图像的颜色信息,即加入颜色不变量,构建彩色描述子;在shape-context算法中改用基于重心点的形状上下文直方图,代替传统的基于各个轮廓点的形状上下文直方图,生成形状上下文描述子。然后把这两种描述子级联成新的联合描述子,依据设定的新的联合距离对特征点进行匹配,得到初始匹配对。最后利用偏最小二乘法消除误匹配,得到精确匹配点对。实验结果表明,提出的算法能够有效提高图像匹配准确率。     

基于二维ISAR图像序列的雷达目标三维重建方法

该文基于刚体目标3维运动过程中的几何不变性,将因子分解法引入ISAR图像处理领域,成功实现了雷达目标从2维到3维的重建。文中基于卡尔曼滤波和最邻近数据关联准则完成不同姿态下2维散射中心的关联,应用因子分解法进行目标结构的分层重建,最后给出评判准则检验重建性能。仿真实验证明,该文方法简单易行,能够有效实现单天线ISAR 3维成像。     

中国城市化水平统计数据的问题分析

城市是一种没有特征尺度的现象,故其只能主观定义,而不能客观界定。城市化水平统计难题因此而生。中国目前的城市化水平至少存在三个方面的问题:一是各地测量标准不一,二是统计上的不适当高估,三是建设中的想当然拔高。由于城市形态和城市规模分布都具有无标度性特征,一个城市的边界和一个区域的城市数目都无法客观确定。这种现象为城市化水平统计过程中的主观操作留下了理论上的漏洞。可是,国内至今没有统一、科学的城市定义,城市化水平统计各地标准参差不齐,水分明显。为了解决上述问题,至少要开展下述基础性工作:基于统计学知识规范城市定义,合理制定城市人口规模门槛,细化城市人口普查单元,在大众中普及正确的城市化概念。     

Dynamic nuclear polarization in crystals of Nd:LaAlO3, a polarized La target for a test of time-reversal invariance

Dynamic nuclear polarization (DNP) has been performed in a magnetic field of 2.35 T at a temperature below 0.3 K in crystals of Nd³⁺ :LaAlO₃. A Q-meter NMR-system was used to measure the polarization of the ²⁷Al and ¹³⁹La nuclei and allowed to characterize the quality of the crystals. The ¹³⁹La polarizations of +47.5% and −56.5% achieved represent an important step in the development of a polarized ¹³⁹La target for a test of time reversal invariance in low-energy neutron reactions.     

Tests for change—points based on recursive u—statistics

Tests for change—points for the location as well as regression models are often based on cumulative sums of recursive residuals. These recursive residuals are also employed in the sequential detection problem. In the context of general estimable parameters (funtionals of the underlying distribution functions), such recursive residuals may be defined in terms of recursive U—statistics. A class of tests for the change—points based on recursive U—statistics has been considered. Along with some invariance principles for recursive U—statistics, asymptotic properties of the proposed tests are studied.     

Structural invariance and age-related performance differences in face cognition.

Perceiving and memorizing faces swiftly and correctly are important social competencies. The organization of these interpersonal abilities and how they change across the life span are still poorly understood. We investigated changes in the mean and covariance structure of face cognition abilities across the adult life span. A sample of 448 participants, with age ranging from 18 to 88 years, completed a battery of 15 face cognition tasks. After establishing a measurement model of face cognition that distinguishes between face perception, face memory, and the speed of face cognition, we used multiple group models and age-weighted measurement models to explore age-related changes. The modeling showed that the loadings and intercepts of all measures are age invariant. The factor means showed substantial decrements with increasing age. Age-related decrements in performance were strongest for the speed of face cognition but were also salient for face perception and face memory. The onset of age decrements is apparent in the 60s for face perception, in the late 40s for face memory, and in the early 30s for speed of face cognition. Implications of these findings on a theoretical and methodological level are discussed, and potential consequences for applied settings are considered. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Personality–cognition relations across adulthood.

Although an increasing number of studies have investigated relations between dimensions of personality and level of cognitive functioning, the research results have been somewhat inconsistent. Furthermore, relatively little is known about whether the personality–cognition relations vary as a function of age in adulthood. The current project examined these issues with data from a sample of 2,317 adults between 18 and 96 years of age who each completed a personality inventory and performed a broad battery of cognitive tests. The results revealed strong relations of the personality trait of Openness with several distinct cognitive abilities and smaller relations of other personality traits with specific cognitive abilities. Comparisons across different age groups indicated that the personality–cognition relations were both qualitatively and quantitatively similar across the adult years. (PsycINFO Database Record (c) 2011 APA, all rights reserved)     

Learning over sets with Recurrent Neural Networks: An empirical categorization of aggregation functions

Numerous applications benefit from parts-based representations resulting in sets of feature vectors. To apply standard machine learning methods, these sets of varying cardinality need to be aggregated into a single fixed-length vector. We have evaluated three common Recurrent Neural Network (RNN) architectures, Elman, Williams & Zipser and Long Short Term Memory networks, on approximating eight aggregation functions of varying complexity. The goal is to establish baseline results showing whether existing RNNs can be applied to learn order invariant aggregation functions. The results indicate that the aggregation functions can be categorized according to whether they entail (a) selection of a subset of elements and/or (b) non-linear operations on the elements. We have found that RNNs can very well learn to approximate aggregation functions requiring either (a) or (b) and those requiring only linear sub functions with very high accuracy. However, the combination of (a) and (b) cannot be learned adequately by these RNN architectures, regardless of size and architecture.     

Planar shape representation and matching under projective transformation

This paper introduces a new representation for planar objects which is invariant to projective transformation. Proposed representation relies on a new shape basis which we refer to as the conic basis. The conic basis takes conic-section coefficients as its dimensions and represents the object as a convex combination of conic-sections. Pairs of conic-sections in this new basis and their projective invariants provides the proposed view invariant representation. We hypothesize that two projectively transformed versions of an object result in the same representation. We show that our hypothesis provides promising recognition performance when we use the nearest neighbor rule to match projectively deformed objects.