A new approach of audio emotion recognition

A new architecture of intelligent audio emotion recognition is proposed in this paper. It fully utilizes both prosodic and spectral features in its design. It has two main paths in parallel and can recognize 6 emotions. Path 1 is designed based on intensive analysis of different prosodic features. Significant prosodic features are identified to differentiate emotions. Path 2 is designed based on research analysis on spectral features. Extraction of Mel-Frequency Cepstral Coefficient (MFCC) feature is then followed by Bi-directional Principle Component Analysis (BDPCA), Linear Discriminant Analysis (LDA) and Radial Basis Function (RBF) neural classification. This path has 3 parallel BDPCA + LDA + RBF sub-paths structure and each handles two emotions. Fusion modules are also proposed for weights assignment and decision making. The performance of the proposed architecture is evaluated on eNTERFACE’05 and RML databases. Simulation results and comparison have revealed good performance of the proposed recognizer.     

大冶硫酸四系列干吸塔结构特点

以干燥塔为例，介绍大冶公司硫酸四系列干吸塔的结构特点。除捕沫层简体及出气管为316L不锈钢外，其余塔体均为碳钢内衬瓷砖结构。塔底为球形底，中心排酸。采用槽管式分酸器，分酸点密度为43个/m^2，设可调闸板以保证各槽酸量均匀；闸板外设置带孔板的分布板，以消除进口酸的冲击，保证槽内酸液面平稳。比较了该塔与常规塔的不同之处，强调了设备节点的优化设计。     

De-oiling contaminated bleaching clay by high-pressure extraction

The disposal and re-use of spent bleaching clay from the vegetable oil processing industry is a problem of growing importance. Although today the only practical way of removal of the spent material is disposal, extraction with organic solvents is a well-known method of de-oiling contaminated bleaching clay. In our investigations we compare the extractibility of two different types of bleaching clays with CO₂ as a solvent. All experiments were carried out with a high-pressure extraction plant. The extraction and separation conditions, temperature and pressure, as well as the CO₂ mass flow, were varied during experiments. The aim of our investigations was a complete separation of the oil from the adsorbent. The latter should then be re-used as bleaching clay. The oil and the bleaching clay were analyzed and tested, respectively. The results show that oil of good quality can be recovered and the bleaching clay still has an activity approximately 50% of fresh clay.     

Solvent extraction of oil from soybean flour II— pilot plant and two-solvent extractions

The process of grinding soybeans to a fine flour and extracting the flour with hexane was studied on a pilot plant scale. The crude oil from the pilot plant study had 15 ppm phosphorus and was suitable for physical refining after a light acid pretreatment and bleaching. The refined oil showed a Lovibond color of 1.4 yellow and 0.3 red. The pilot plant study also showed that grinding of the soybeans and the separation of solid from miscella were the most difficult steps in solvent extraction with fine flour. A laboratory study on separation of miscella from meal by aqueous ethanol reduced the hold-up volume, but it did not remove all the miscella. A test with betacarotene showed that only the miscella outside the flour particles was displaced. Aqueous ethanol solutions used as a second solvent extracted additional nontriglyceride materials (primarily phospholipids) from the meal. Also, the free fatty acid content of the oil was increased with aqueous ethanol solution wash. The quality of the extracted crude oil was lowered by using a second solvent, but it had the advantage of needing only one centrifugation to separate miscella from meal.     

Effect of sample size on solvent extraction for detecting cocontinuity in polymer blends

The effect of sample size on the results of solvent extraction measurements for detecting cocontinuity in polymer blends was investigated. Poly(ethylene oxide)/polystyrene (PEO/PS) blend samples of several thicknesses were analyzed by removing the PEO phase using water extraction. The experimental degree of continuity was shown to have a linear dependence on the reciprocal of sample thickness. A model is proposed to explain this dependence and to allow the bulk or true degree of continuity to be determined. Measurement of the bulk degree of continuity is useful for understanding properties of cocontinuous polymer blends such as electrical conductivity, impact strength, or tensile strength.     

基于多特征融合的运动物体识别方法研究

为了实现运动目标的实时检测,本文基于多特征融合,提出了一种运动物体分级识别方法.首先通过对运动目标的分割,分析得到各个物体的形状矩特征、面积大小及运动速度.对上述特征,采用支持向量机方法对其矩特征进行快速的一级分类,然后对一级分类结果及各物体的面积、速度等信源根据其对分类的重要程度,采用模糊积分技术进行多特征融合分类.实验表明,本方法具有较强的学习能力及较高的识别精度,可有效地应用于道路监控等系统.     

基于KPCA和SVM的火箭发动机试验台故障诊断方法

为了解决液体火箭发动机试验台的故障诊断问题,提出了一种基于核主元分析(KPCA)特征提取和支持向量多分类机(SVM)的故障诊断方法,该方法首先利用核主元分析对试验台标准故障样本进行特征提取,通过特征分析,建立适合于试验台故障状态识别的层次多分类支持向量机,并对其进行训练,然后将试验数据在主元上投影,输入到训练好的支持向量多分类器,对试验台故障状态进行识别.该方法充分利用了核主元分析强大的非线性特征提取能力和支持向量分类机良好的小样本泛化特性,解决了试验台故障诊断中的小样本、非线性模式识别问题.对试验台的试验结果表明,该方法是有效的、可行的.     

基于HIS颜色空间颜色对的图像检索系统

提出了一种在HIS空间中进行的基于颜色对的图像检索算法.为突出颜色在画面中的位置在内容检索上的作用,算法在HIS颜色空间的上对各颜色分量进行量化,用颜色对来描述图像的颜色特征,定义了基于颜色对的相似度度量,利用图像颜色特征进行相似度匹配和检索.实验结果表明,该算法能比较快速、有效地针对图像内容进行检索.     

红外与可见光复合寻的制导中的快速图像配准方法

对现有多传感器图像配准方法在红外与可见光复合寻的制导应用中存在的计算量大等不足,提出了一种基于部分Hausdorff距离的快速方法和一种两集合相似度的定义,从而实现了有效减小误匹配特征点对的红外与可见光图像快速配准方法.此算法通过分别提取两幅图像边缘上显著特征点,再根据提出的快速方法和两集合相似度定义,粗略估计出图像间的旋转角度、平移量;然后依据此估计值,缩小特征点匹配时匹配特征点搜索范围,求得一致特征点对;最后用最小二乘法求解最优变换参数.实验结果表明了该方法的有效性、快速性.     

一种基于步幅长度及频域特征的步态识别方法

步态识别作为一种较新的生物认证技术有其独特优势,它旨在根据人行进过程中的行走步态模式来识别其身份.提出一种基于步幅长度及频域特征的步态识别方法:首先采用背景减除算法实现运动人体检测,并根据步幅长度变化特征进行周期分割.然后提取步幅长度极值结合关键帧的频幅均值构成五维特征向量,最后用标准模式分类器在中科院自动化所提供的CASIA步态数据库中进行实验,在降低了算法复杂度的同时,获得了令人鼓舞的实验结果.