A method of Biomimetic Pattern Recognition for face recognition

A new method of face recognition, based on Biomimetic Pattern Recognition and Multi-Weights Neuron Network, had been proposed. A model for face recognition that is based on Biomimetic Pattern Recognition had been discussed, and a new method of facial feature extraction also had been introduced. The results of experiments with BPR and K-Nearest Neighbor Rules showed that the method based on BPR can eliminate the error recognition of the samples of the types that not be trained, the correct rate is also enhanced.

一种基于粒子的牛顿流体与粘弹性流体统一模拟方法

在越来越受到人们关注的基于物理流体动画领域,目前分别模拟牛顿流体或粘弹性流体的方法很多,但很少有统一模拟两者的方法.文中基于光滑粒子流体动力学方法,通过对传统纳维-斯托克斯方程添加弹性应力项,提出了一种新的统一模拟牛顿流体和粘弹性流体的方法.通过实验说明该方法不仅有效,易于实现,而且具有良好的可控性,仅仅通过调节参数就可以模拟不同粘弹性、不同类型的流体现象.

Investigation on molecular and crystal structures of metal complexes with aminopolycarboxylic acids .1. Synthesis and structure of Na-2[Fe-III(ida)(2)](2)center dot 3H(2)O

The crystal structure of the title complex salt has been determined by single-crystal X-ray structure analysis. The crystal data areas follows; Monoclinic, P2(1)/c, a=15.6480(10)Angstrom, b=16.7870(10)Angstrom, c=10.347(2)Angstrom, beta=90.790(10), V=2717.7(6)Angstrom(3), Z=3, and R=0.0333 for 4789 unique reflections. The complex anion has a pseudo-octahedral structure distorted more than the Cr-III and Co-III analogs, in which each, iminodiacetato ligand (ida(2-)) is coordinated in a facial fashion with the two N atoms in a cis configuration, resulting in an unsym-fac structure.

Second-order solutions for random interfacial waves in N-layer density-stratified fluid with steady uniform currents

In the present paper, the random inter facial waves in N-layer density-stratified fluids moving at different steady uniform speeds are researched by using an expansion technique, and the second-order a symptotic solutions of the random displacements of the density interfaces and the associated velocity potentials in N-layer fluid are presented based on the small amplitude wave theory. The obtained results indicate that the wave-wave second-order nonlinear interactions of the wave components and the second-order nonlinear interactions between the waves and currents are described. As expected, the solutions include those derived by Chen (2006) as a special case where the steady uniform currents of the N-layer fluids are taken as zero, and the solutions also reduce to those obtained by Song (2005) for second-order solutions for random interfacial waves with steady uniform currents if N=2.

改进的统计不相关最优鉴别矢量集

该文对统计不相关最优鉴别矢量集算法进行研究,在分析统计不相关最优鉴别矢量集算法的基础上提出了一种改进的方法。该方法在类内散布矩阵的特征空间中求解统计不相关最优鉴别矢量集。为了加快特征抽取速度,利用基于图像鉴别分析的维数压缩方法,对图像数据进行了压缩。在ORL和Yale人脸数据库的数值实验,验证本文所提出的方法的有效性。

基于模糊神经网络的人脸识别方法研究

人脸识别是模式识别研究领域的重要课题,具有广阔的应用前景。本文提出了基于模糊神 经网络的人脸识别方法。首先用最优鉴别分析方法提取人脸的最优鉴别矢量集,构成特征空间,然后在 特征空间中设计模糊神经网络分类器。在ORL人脸图象库上的实验结果表明了该方法的有效性。

统计不相关最佳鉴别矢量集的本质研究

对统计不相关最佳鉴别矢量集的本质进行研究 ,在基于总体散布矩阵特征分解的基础上 ,构造了一种白化变换 ,使得变换后的样本空间中的总体散布矩阵为单位矩阵 ,这样使得传统的最佳鉴别矢量集算法得到的均是具有统计不相关的最佳鉴别矢量集 ,从而揭示了统计不相关最佳鉴别变换的本质———白化变换加普通的线性鉴别变换。该方法的最大优点在于所获得的最优鉴别矢量同时具有正交性和统计不相关性。该方法对代数特征抽取具有普遍适用性。用ORL人脸数据库的数值实验 ,验证了该方法的有效性

基于扰动方法的广义最佳鉴别矢量集求解的一种迭代算法

本文对广义最佳鉴别矢量的求解方法进行研究 ,根据矩阵的扰动理论 ,改进了郭提出的求解广义最佳鉴别矢量的一种迭代算法 ,提出了求解广义最佳鉴别矢量的一种新的迭代算法 .本文算法的一个突出优点是随着类别数目的增加 ,计算时间反而缩短 ;而老算法随着类别数目的增加计算时间随着增加 ;不仅如此 ,新算法的识别率不劣于老算法 .在 ORL人脸数据库的数值实验 ,验证了上述论断的正确性

工程地质三维建模与可视化技术应用开发

Synthetic Geology Information System (SGIS) is an important constituent part of the theory of Engineering Geomechanics Mate-Synthetic (EGMS), and is the information system more suited for the collection, storage, management, analysis and processing to the information coming from engineering geology,' geological engineering and geotechnical engineering. Its contents involve various works and methods of the investigation, design, and construction in different stages of the geological engineering. Engineering geological and three-dimensional modeling and visualization is the fundamental part of the SGIS, and is a theory, method and technique by which, adopting the computer graphics and image processing techniques, the data derived from engineering geological survey and the calculated results obtained from the geomechanical numerical simulation and analysis are converted to the graphics and images displayed on the computer screen and can be processed interactively. In this paper, the significance and realizing approaches of the three-dimensional modeling and visualization for the complex geological mass in the engineering geology are discussed and the methods of taking advantage of the interpolation and fitting for the scattered and field-surveyed data to simulate the geological layers, such as the topography and earth surface, the groundwater table and the stratum boundary, are researched into. At the mean time, in mind the characteristics of the structure of the basic data for three-dimensional modeling, its visual management can be resolved into the engineering surveyed database management module, plot parameter management module and data output module and the requirement for basic data management can be fulfilled. In the paper, the establishment and development of the three-dimensional geological information system are probed tentatively, and an instance of three-dimensional visual Engineering Distribution Information System (EDIS), theConstruction Management Information System for an airport, in which the functions, such as the real-time browse among the three-dimensional virtual-reality landscapes of the airport construction from start to finish, the information query to the airport facility and the building in the housing district and the recording and playback of the animation sets for the browse and the takeoff and landing of the planes, is developed by applying the component-mode three-dimensional virtual-reality geological information system (GIS) software development kits (SDK), so the three-dimensional visual management platform is provided for the airport construction. Moreover, in the gaper, integrated with the three-dimensional topography visualization and its application in the Sichuan-Tibet Highways, the method of the digital elevation model (DEM) data collection from the topographic maps is described, and the three-dimensional visualization and the roaming about the terrain along the highway are achieved through computer language programming. Understanding to the important role played by the varied and unique topographical condition in the gestation and germination of the highly-dense, frequently-arising and severely-endangered geological hazards can be deepened.

动漫行业创业者胜任特征研究

Venture has become an important force for the sustainable economic growth of China, However, research on entrepreneurs is far failed to meet the actual needs, Animation and Comic Industry is an important component of the state-led Cultural and Creative Industries. Entrepreneurship is the rapid growth of population, but the empirical research on the Animation and Comic Industry entrepreneurs’ Competent Characteristic Model is rarely heard. As a manager of a national incubator which major in Cultural and Creative Industry, study these entrepreneurs competent characteristics is my responsibility. As an exploration, the study uses Competent Characteristic Model of classical design methods and behavior events interview method, and selects entrepreneurs which of animation and comic industry business start-up period as the objects. Through coding analysis of 31 entrepreneurs’ interview information, completed the amendment to the dictionary coding, and ultimately construct the animation and comic industry entrepreneurs need to have the threshold competence model and differentiating competence model. Differentiating competence model includes six competencies: tolerance and optimism, cooperation, creative innovation, pool talents, vision, decision-making and tempo control. Threshold competence model includes twenty features, divided into three dimensions: personal characteristics, knowledge quality, capacity of the quality. Knowledge and capacity of the quality are closely related to the core work of entrepreneurs, Personal characteristics concern the deep-seated motivation of the entrepreneurs. The study also analyses and discusses the characteristics of the entrepreneurs on the demographic variables.

面孔身份与面部表情识别中的区分度效应

Whether facial identity and facial expression was processed independently has long been a controversy. Studies at levels of experimental, neuropsychological, functional imaging and cell-recording all failed to consistently support either independent or interdependent processing. Present study proposed that familiarity and discriminability of facial identity and expression was important variable in mediating the relation between facial identity and facial expression recognition. Effects of familiarity on recognition of facial identity and expression had been examined (e.g. Ganel & Goshen-Gottstein, 2004) but the role of the discriminability in recognition of facial identity and expression has not yet been carefully examined. To examine the role of discriminability of facial identity and expression, 8 experiments were conducted with Garner’s speeded classification task in recognition of identity and expression of unfamiliar faces. The discriminability of facial identity and expression was manipulated, and the measurements of Garner interference and facilitation indicated that: 1. The discriminability of facial identity and expression mediate the relation between facial identity and expression recognition. Four possible discriminability combinations between identity and expression predicted 4 interference patterns between them. Low discriminability accounted for the interference either in facial identity judgment or in facial expression judgment task. 2. The measurements of eye movements indicated that either in facial identity or in facial expression recognition low discriminability led to a narrowly-distributed eye fixation pattern while high discriminability led to a widely-distributed eye fixation pattern. 3. By combining the morphing technique with the Garner paradigm, study 2 successfully demonstrated the linar relation between discriminability and Garner facilitation effects, confirmed the discriminability effects in the measurements of Garner facilitation effects.. 4. By providing the varying information of facial expression, study 2 revealed that varying information improved the discriminability of facial expression, and then enhanced the recognition of facial expression. All the results indicated that the discriminability of facial identity and expression could mediate the independent or interdependent processing between them, and the discriminability effects on recognition of identity and expression of unfamiliar faces was identified. The results from interference effects and facilitation effects both indicated that the dimensional relation between facial identity and expression was separable but not asymmetric claimed by previous studies（Schweinberger et al, 1998, 1999）. Absolutedly independent or interdependent processing between facial identity and expression recognition were both impossible, discriminability of identity and expression mediated the relation between them. The discriminability effects revealed in present study could explain the conflicts between existing findings well.

精神分裂症患者及其未患病一级亲属的面孔情绪知觉研究

Schizophrenia is a heritable disorder. However, molecular genetics and related research area have not unmasked the nature and mechanisms of this disorder. Therefore, many researchers begin to explore the pathology mechanism from other approaches. High-risk study is one of the promising approaches. In this study, we mainly focused on facial emotion perception in schizophrenia and their non-psychotic first-degree relatives, and attempted to explore whether facial emotion perception is the potential biological marker of schizophrenia. This dissertation comprises 4 studies. In the first study, we conducted a meta-analysis on behavioral data of facial emotion perception in schizophrenia. Our findings showed that patients demonstrated general deficits in both facial emotion perception and facial processing tasks. In the second study, sixty-nine patients with schizophrenia and 56 of their first-degree relatives (33 parents and 23 siblings), and 92 healthy controls (67 younger and 25 older healthy controls) completed a set of facial emotion perception tasks. The results validated that patients with schizophrenia displayed general deficits in facial emotion perception. Study two also demonstrated that siblings of patients performed equally well compared to the corresponding younger healthy controls in all the facial emotion perception tasks, while the parents of patients behaved significantly worse than the corresponding older healthy controls in the composite index of facial emotion perception tasks. The results suggest that relatives of patients display more severely declining in facial emotion perception with the increasing of age. In the third study, we used an automated voxel-wise technique, activation likelihood estimation (ALE) to provide an objective, quantitative evaluation of facial emotion processing in schizophrenia. Our findings demonstrated a marked under-recruitment of the amygdala, accompanied by a substantial limitation in activation in schizophrenia throughout a ventral temporal-basal ganglia-prefrontal cortex ‘social-brain’ system may be central to the difficulties patients experience when processing facial emotion. In the last study, we did an fMRI study about facial emotion perception in 12 patients with schizophrenia, 12 non-psychotic siblings of patients and 12 healthy controls. The results suggest that siblings of patients demonstrate abnormal activation in a variety of brain areas, including prefrontal gyrus, insula, parahippocampal gyrus and superior temporal gyrus. Taken together, the current findings suggest facial emotion perception may be a potential biological marker of schizophrenia.

Towards an Example-Based Image Compression Architecture for Video-Conferencing

This paper consists of two major parts. First, we present the outline of a simple approach to very-low bandwidth video-conferencing system relying on an example-based hierarchical image compression scheme. In particular, we discuss the use of example images as a model, the number of required examples, faces as a class of semi-rigid objects, a hierarchical model based on decomposition into different time-scales, and the decomposition of face images into patches of interest. In the second part, we present several algorithms for image processing and animation as well as experimental evaluations. Among the original contributions of this paper is an automatic algorithm for pose estimation and normalization. We also review and compare different algorithms for finding the nearest neighbors in a database for a new input as well as a generalized algorithm for blending patches of interest in order to synthesize new images. Finally, we outline the possible integration of several algorithms to illustrate a simple model-based video-conference system.

Estimation of Pose and Illuminant Direction for Face Processing

In this paper three problems related to the analysis of facial images are addressed: the illuminant direction, the compensation of illumination effects and, finally, the recovery of the pose of the face, restricted to in-depth rotations. The solutions proposed for these problems rely on the use of computer graphics techniques to provide images of faces under different illumination and pose, starting from a database of frontal views under frontal illumination.