A weighted feature reduction method for power spectra of radar HRRPs

Feature reduction is a key process in pattern recognition. This paper deals with the feature reduction methods for a time-shift invariant feature, power spectrum, in Radar Automatic Target Recognition （RATR） using High-Resolution Range Profiles （HRRPs）. Several existing feature reduction methods in pattern recognition are analyzed, and a weighted feature reduction method based on Fisher＇s Discriminant Ratio （FDR） is proposed in this paper. According to the characteristics of radar HRRP target recognition, this proposed method searches the optimal weight vector for power spectra of HRRPs by means of an iterative algorithm, and thus reduces feature dimensionality. Compared with the method of using raw power spectra and some existing feature reduction methods, the weighted feature reduction method can not only reduce feature dimen- sionality, but also improve recognition performance with low computation complexity. In the recognition experiments based on measured data, the proposed method is robust to different test data and achieves good recognition results.     

Multi-manifold-based skin classifier on feature space Voronoï regions for skin segmentation

Skin segmentation is a crucial and a challenging step in many face and gesture recognition techniques and it has various applications in human computer interaction, objectionable content filtering, image retrieval and many more. In this article, we propose a novel skin segmentation method, which uses multi-manifold-based skin classification of feature space skin candidate Voronoï regions to achieve accurate skin segmentation. The state-of-the-art skin segmentation techniques reported in this article focus on discrimination between textural feature vectors belonging to skin and non-skin classes. In contrast, the proposed method focuses on discrimination between textural feature vectors belonging to skin and skin-like (non-skin) classes, which lead to higher skin classification accuracy. Furthermore, we introduce a novel image segmentation technique based on spatial and feature space Dirichlet tessellation (also called a Voronoï diagram) to achieve feature space segmentation of skin candidate regions of an image. These feature space segments will then be classified using a multi-manifold-based skin classifier. The proposed skin segmentation method was evaluated on two benchmark skin segmentation data sets and its results were compared with four other state-of-the-art methods proposed for skin segmentation. The experimental results reported in this article confirm that the proposed method outperforms the existing skin segmentation approaches in terms of false alarm rates in the skin segmentation process. Also, the proposed method results in the lowest minimal detection error compared to the existing methods reported in this article.     

基于稀疏约束的SAR目标特征提取方法研究

特征提取是合成孔径雷达目标识别关键技术与核心任务。为了更好地提取目标特征,稀疏约束将被添加在非负矩阵分解法中,并应用于图像目标特征提取,通过利用稀疏约束的非负矩阵分解方法对sAR目标图像进行分解,构建具有稀疏性的目标特征矢量,提高了特征矢量的类内相似性与类间差异性。利用基于支持向量机的分类方法对MSTAR数据进行目标识别试验,试验结果表明,添加稀疏约束的NMF方法与PCA、ICA以及一般NMF特征提取方法相比,能够显著提高目标识别的稳定性和准确率。     

基于正则化零空间鉴别分析的人脸识别

提取有效的特征对高维数据的模式分类起着关键 作用,针对现有故障特征维数过高的问题,本文提 出了一种基于正则化零空间线性鉴别分析(Exponential Regularized Null Space Linear Discriminant Analysis, ERNSLDA)的特征提取方法。零空间线性判别分析已经在数据降维和特征提取上展现出良好 的性能,在 本文中,首先对类内样本矩阵进行正则化处理,避免小样本问题,其次对判别准则进行指数 化处理。所提 方法集成了NSLDA和RLDA在模式识别上的优势,有效地提高了人脸识别的精度,在ORL和YALE 数据库上的仿真实验证了本文所提方法的有效性。     

改进CSP算法的联合特征优化

传统的公共空间模式分解需要大量输入通道、缺乏频域信息,文章分别从改进CSP滤波器、构建关于CSP的联合特征、优化识别过程三个方面完善CSP算法的不足。首先,提出基于S变换的公共空间滤波器成分选择算法--CSPS。并将CSPS与EMD、EEMD、双谱分析结合,构建EMD-CSPS、EEMD-CSPS、双谱-CSPS三种联合特征并比较判别效果。最后,使用优化后的联合特征,一方面,对支向量机惩罚因子和内核参数进行优化,确定惩罚因子最优取值范围和最具分类稳定性的内核函数;另一方面,分别采用支持向量机和线性判别分析进行特征识别与比较。文章设计了左右手想象运动思维任务实验,获取实验数据集,并结合BCI竞赛数据集,从分类正确率和响应时间两个指标出发,分析各优化方法有效性。结果表明:采用S变换优化后的双谱-CSPS特征在LDA分类器下,获得较高的分类正确率和较低的系统建模时间。      

Low Latency Fog-Centric Deduplication Approach to Reduce IoT Healthcare Data Redundancy

In today’s highly computerized society, detection and recognition of text present in natural scene images is complex and difficult to be properly recognized by human vision. Most of the existing algorithms and models mainly focus on detection and recognition of text from still images. Many of the recent machine translation systems are built using the Encoder-Decoder framework which works on the format of encoding the sequence of input and then based on the encoded input, the output is decoded. Both the encoder and the decoder use an attention mechanism as an interface, making the model complex. Aiming at this situation, an alternative method for recognition of texts from videos is proposed. The proposed approach is based on a single Two-Dimensional Convolutional Neural Network (2D CNN). An algorithm for extracting features from an image called the crosswise feature extraction is also proposed. The proposed model is tested and shows that crosswise feature extraction gives better recognition accuracy by requiring a lesser period of time for training than the conventional feature extraction technique used by CNN.

     

一种基于规范互信息和动态冗余信号识别技术的特征选择方法

为了实现对模式识别、信号处理等领域中数据的有效表达,提出了一种基于规范互信息和动态冗余信号识别技术的特征选择方法。该方法采用规范互信息对特征相关性和冗余性进行测量,并通过一种动态冗余信号识别技术在特征全集中进行冗余特征的筛选。分类实验结果表明所提特征选择方法性能优于典型的特征选择方法。     

采用加权协方差矩阵描述子的OCR识别方法

传统光学字符识别(Optical Character Recognition,OCR)方法一般只提取图像亮度特征,在图像退化较严重时识别准确率不高。针对这一问题,提出一种新的扫描字符特征提取方法。除各通道亮度外,还提取像素位置、亮度的一阶导、二阶导等特征构成特征图像,并根据各个特征对图像的贡献程度进行加权处理。计算以当前像素为中心的局部区域特征图像块的协方差矩阵作为当前像素的描述子,然后在黎曼空间对字符实施分类。实验结果表明,采用典型的结构化分类器时,该特征提取方法对字符识别的准确率高于传统方法,表现出较强的鲁棒性。     

基于双通道卷积神经网络的视频目标移除取证算法

针对现有数字视频目标移除取证算法的伪造帧识别准确率低的问题,本文提出了一种基于双通道卷积神经网络的视频目标移除取证算法。该算法利用双通道结构,分别提取视频绝对帧差图像的RGB特征和噪声特征,并利用双线性池化对二者进行特征融合,而后通过分类层输出视频帧的分类结果,从而有效地识别经过篡改的视频帧。其中,RGB通道能够发现绝对帧差图像中不自然的篡改边界和对比度,噪声通道能够发现原始区域和篡改区域之间噪声的不一致性。此外,算法在网络前端增加了预处理层来放大篡改视频帧的伪造痕迹。实验结果显示,所提算法有效地提高了伪造视频帧的识别准确率,且相对于传统的单通道网络结构,双通道特征融合的方式取得了更好的检测性能。      

改进非负矩阵分解的神经网络人脸识别

为了提高人脸识别正确率,提出基于改进非负矩阵分解的神经网络人脸识别算法。首先利用改进的非负矩阵分解对人脸图像进行特征提取,提高非负矩阵分解速度。接着将提取出的特征信息作为神经网络学习入口进行特征训练,由于神经网络在学习过程中,容易出现局部最小值且收敛速度慢等问题,为此采用改进的遗传算法对神经网络进行优化处理,获得最终的人脸识别结果。实验结果表明:利用改进的非负矩阵分解方法能够降低神经网络的分类训练负荷量和运算量,提高人脸识别识别率。通过和各种方法比较可知,本方法的人脸识别率都较高。本方法人脸特征分解速度快,提高了神经网络训练前期精度和收敛速度,使得人脸识别正确率高。当特征向量个数达到40以上时,人脸识别正确率保持95%以上。     

A Supervised Feature-Projection-Based Real-Time EMG Pattern Recognition for Multifunction Myoelectric Hand Control

Electromyographic (EMG) pattern recognition is essential for the control of a multifunction myoelectric hand. The main goal of this study was to develop an efficient feature- projection method for EMG pattern recognition. To this end, a linear supervised feature projection is proposed that utilizes a linear discriminant analysis (LDA). First, a wavelet packet transform (WPT) is performed to extract a feature vector from four-channel EMG signals. To dimensionally reduce and cluster the WPT features, an LDA, then, incorporates class information into the learning procedure, and identifies a linear matrix to maximize the class separability for the projected features. Finally, a multilayer perceptron classifies the LDA-reduced features into nine hand motions. To evaluate the performance of the LDA for WPT features, the LDA is compared with three other feature-projection methods. From a visualization and quantitative comparison, it is shown that the LDA produces a better performance for the class separability, plus the LDA-projected features improve the classification accuracy with a short processing time. A real-time pattern-recognition system is then implemented for a multifunction myoelectric hand. Experiments show that the proposed method achieves a 97.4% recognition accuracy, and all processes, including the generation of control commands for the myoelectric hand, are completed within 97 ms. Consequently, these results confirm that the proposed method is applicable to real-time EMG pattern recognition for multifunction myoelectric hand control.     

An Exhaustive Multi Factor Face Authentication Using Neuro-Fuzzy Approach

The face authentication is a challenging task to validate the user with uncontrolled environment like variations on expression, pose, illumination and occlusion. In order to address these issues, the proposed work provides solution by considering all these factors in inter and intra personal face authentication. During enrollment process, the facial region of still image for the authorized user is detected and features are extracted using local tetra pattern (LTrP) technique. The features are given as input to the neural network namely fuzzy adaptive learning control network (FALCON) for training and classification of features. During authentication process, an image that can vary with expression, pose, illumination and occlusion factors is taken as test image and the test image is applied with LTrP and FALCON to train the features of test image. Then, these trained features are compared with existing feature set by using new proposed multi factor face authentication algorithm to authenticate a person. This work is evaluated among 1150 face images which are collected from JAFFE, Yale, ORL and AR datasets. The overall performance of the work is evaluated by authenticating 1106 images from 1150 constrained images. The second phase of the research work finally produces highest recognition rate of 96% among conventional methods.

     

Action recognition based on statistical analysis from clustered flow vectors

A new technique for action clustering-based human action representation on the basis of optical flow analysis and random sample consensus (RANSAC) method is proposed in this paper. The apparent motion of the human subject with respect to the background is detected and localized by using optical flow analysis. The next task is to characterize the action through the frequent movement of the optical flow points or interest points at different regions of the moving subject. The RANSAC algorithm is used to filter out any unwanted interested points all around the scene and keep only those that are related to that particular subject’s motion. From the remaining salient key interest points, the area of the human body within the frame is estimated. The rectangular area surrounding the human body is then segmented both horizontally and vertically. Now, the percentage of change of interest points in each horizontal and vertical segments from frame to frame is estimated. Similar results are obtained for different persons performing the same action and the corresponding values are averaged for respective segments. The matrix constructed by this strategy is used as a feature vector for that particular action. Similar data are calculated for each block created at the intersections of the horizontal and vertical segments. In addition to these, the change in the position of the person along X- and Y-axes is accumulated for an action and included in the feature vectors. Afterward, for the purpose of recognition using the extracted feature vectors, a distance-based similarity measure and a support vector machine-based classifiers have been exploited. Several combination of the feature vectors is examined. From extensive experimentation upon benchmark motion databases, it is found that the proposed method offers not only a very high degree of accuracy but also computational savings.     

采用新型纹理特征的2DLDA人脸识别算法

针对现有基于纹理特征的人脸识别算法中纹理特征维数偏大且对噪声较敏感等不足,提出了用于描述人脸图像大尺度局部特征的中心四点二元模式（Center Quad Binary Pattern, C-QBP）和用于描述图像小尺度局部特征的简化四点二元模式（Simplified Quad Binary Pattern, S-QBP）两种互补的新型纹理特征。在此基础上,实现基于新型纹理特征的2DLDA人脸识别算法。首先对人脸图像进行多级分割,再对所产生的图像块提取C-QBP和S-QBP纹理特征,构建纹理特征矩阵。最后,采用2DLDA子空间学习算法实现基于新型纹理特征的人脸识别。实验结果表明,本文所提出的人脸识别算法的识别率明显高于其他基于纹理特征和子空间学习的人脸识别算法。当每一类训练样本数统一设置为5,特征维数为48×4时,在ORL人脸库上,本文所提出的人脸识别算法的识别率达98.68%;在YALE人脸库上,特征维数为48×36时,识别率达99.42%;在FERET人脸库上,特征维数为48×26时,识别率为91.73%。      

改进ASM和分类树表情识别算法

在现有的人脸表情识别系统中,速度和识别率是最重要的两个衡量标准,为提高人脸表情判别速度和识别率,采用了一种改进了的ASM和分类树表情识别的新方法。首先对传统的ASM的特征点定位过程进行改进,主要用条带法进行局部特征点定位和使用选择性特征点提取算法来提高特征点定位的速度和准确性。用分类树识别算法来改进经典的模板匹配分类法。实验结果表明,在JAFFE人脸表情数据库中进行实验可以获得更好的识别效果。     

Variability Analysis of SBOX With CMOS 45 nm Technology

An efficient sign language recognition system (SLRS) can recognize the gestures of sign language to ease the communication between the signer and non-signer community. In this paper, a computer-vision based SLRS using a deep learning technique has been proposed. This study has primary three contributions: first, a large dataset of Indian sign language (ISL) has been created using 65 different users in an uncontrolled environment. Second, the intra-class variance in dataset has been increased using augmentation to improve the generalization ability of the proposed work. Three additional copies for each training image are generated in this paper, by using three different affine transformations. Third, a novel and robust model using Convolutional Neural Network (CNN) have been proposed for the feature extraction and classification of ISL gestures. The performance of this method is evaluated on a self-collected ISL dataset and publicly available dataset of ASL. For this total of three datasets have been used and the achieved accuracy is 92.43, 88.01, and 99.52%. The efficiency of this method has been also evaluated in terms of precision, recall, f-score, and time consumed by the system. The results indicate that the proposed method shows encouraging performance compared with existing work.

     

Spatiotemporal emotion recognition based on 3D time-frequency domain feature matrix

The research of emotion recognition based on electroencephalogram (EEG) signals often ignores the relatedinformation between the brain electrode channels and the contextual emotional information existing in EEG signals,which may contain important characteristics related to emotional states. Aiming at the above defects, aspatiotemporal emotion recognition method based on a 3-dimensional (3D) time-frequency domain feature matrixwas proposed. Specifically, the extracted time-frequency domain EEG features are first expressed as a 3D matrixformat according to the actual position of the cerebral cortex. Then, the input 3D matrix is processed successivelyby multivariate convolutional neural network (MVCNN) and long short-term memory (LSTM) to classify theemotional state. Spatiotemporal emotion recognition method is evaluated on the DEAP data set, and achievedaccuracy of 87.58% and 88.50% on arousal and valence dimensions respectively in binary classification tasks, aswell as obtained accuracy of 84.58% in four class classification tasks. The experimental results show that 3D matrixrepresentation can represent emotional information more reasonably than two-dimensional (2D). In addition,MVCNN and LSTM can utilize the spatial information of the electrode channels and the temporal context information of the EEG signal respectively.     

一种基于WRCE-CSP的运动想象电位识别方法

针对脑一机接口系统中运动想象脑电信号（Electroencephalography,EEG）的模式识别问题,提出了加权节律成分提取（WeightedRhythmicComponentExtraction,WRCE）与共空间模式（CommonSpacePattern,CSP）相结合的特征提取方法,并使用Fisher线性判别分析进行分类。采用2003年的BCI竞赛数据Datasetm对该方法进行评估,测试数据的分类正确率达到86．13％,比使用传统CSP方法进行特征提取时的分类正确率提高了5．71％,表明该方法可有效地应用于运动想象EEG的模式识别中。     

Gait recognition using Pose Kinematics and Pose Energy Image

Many of the existing gait recognition approaches represent a gait cycle using a single 2D image called Gait Energy Image (GEI) or its variants. Since these methods suffer from lack of dynamic information, we model a gait cycle using a chain of key poses and extract a novel feature called Pose Energy Image (PEI). PEI is the average image of all the silhouettes in a key pose state of a gait cycle. By increasing the resolution of gait representation, more detailed dynamic information can be captured. However, processing speed and space requirement are higher for PEI than the conventional GEI methods. To overcome this shortcoming, another novel feature named as Pose Kinematics is introduced, which represents the percentage of time spent in each key pose state over a gait cycle. Although the Pose Kinematics based method is fast, its accuracy is not very high. A hierarchical method for combining these two features is, therefore, proposed. At first, Pose Kinematics is applied to select a set of most probable classes. Then, PEI is used on these selected classes to get the final classification. Experimental results on CMU's Mobo and USF's HumanID data set show that the proposed approach outperforms existing approaches.