Video shot motion characterization based on hierarchical overlapped growing neural gas networks

Automatic video shot motion characterization is an important step in video indexing and retrieval after temporal video segmentation. This paper describes a hierarchical overlapped architecture (HOGNG) based upon the growing neural gas (GNG) network [7] to perform this task. The proposed architecture combines the unsupervised and supervised learning schemes in GNG. As higher-level GNGs overlap, the final classification is obtained by fusing the individual classifications generated by the top-level overlapping GNGs. In addition, we employ prefiltering and postfiltering for improving the classification accuracy. Experimental results are presented to show the good classification accuracy of the proposed algorithm on real MPEG video sequences.P.N. Suganthan: Correspondence to     

Iris sensor identification in multi-camera environment

Large-scale identity projects such as the Unique Identification Authority of India (UIDAI) comprise of multiple individual organizations, which may use different sensors for enrolling the individuals while the data obtained at the time of verification can be collected from a different sensor. In such multi-camera scenario, it is imperative to perform image-based iris sensor identification. In this research, we propose an efficient algorithm to identify the sensor from which the iris image is captured. The proposed algorithm is the amalgamation of SVM fitness function based Bacteria Foraging (BF) feature selection and fusion of multiple features such as Block Image Statistical Measure (BISM), High Order Wavelet Entropy (HOWE), Texture Measure (TM), Single-level Multi-orientation Wavelet Texture (SlMoWT), and Image Quality Measures (IQM). The selected features are then given input to a supervised classification algorithm for iris sensor identification. The second contribution of this research is developing two sets of multisensor iris image databases that, in total, contain 6000 images with over 150 subjects. The results show that the proposed sensor classification algorithm is computationally very fast and yields an accuracy of over 99% on multiple databases.     

Hybrid Control for Autonomous Mobile Robot Navigation Using Neural Network Based Behavior Modules and Environment Classification

A hybrid control architecture combining behavior based reactive navigation and model based environment classification has been developed. It is also hybrid in the sense that both competitive coordination and cooperative coordination are used for the BBC (Behavior Based Control) part. The contributions are as follows. First, a Neural Network (NN) in charge of environment classification has been developed based on 16 prototypes of topological maps roughly describing various local navigation environments. This environment classification NN not only enables the navigator to avoid local minimum points but also eliminates the requirement for prior detailed modeling of the environment since it needs to memorize only rough information on local environments encountered along the way that might be sufficient for navigation. Next, an NN based reactive behavior controller will be trained to learn human steering commands for each of the 16 prototype local environments. Third, the modified potential field (MPF) method obtained by adding the free space vector as the third component is used to select a particular reactive behavior in conjunction with the classification NN. Finally, a hybrid control architecture integrating all three of these concepts was developed. It avoids local minimum traps as well as solves the problems of poor obstacle clearance or oscillation. It is robust against sensor noise and adaptive to dynamic environments. This hybrid architecture is also amenable to easy addition of new behaviors due to the modularity of the BBC architecture. The effectiveness of the proposed architecture has been verified through both computer simulation and an actual robot called MORIS (MObile Robot as an Intelligent System).     

A multigradient algorithm using a mixture of experts architecture for land cover classification of multisensor images

An algorithm for supervised classification of multisensor images is proposed. The mixture of experts (ME) architecture with dynamic weight allocation is used for multiclass classification. Here the classification is treated as a maximum likelihood problem and the synaptic weights of the expert network and gating network are updated by a stochastic multigradient approach. Data from an optical sensor with four bands and a synthetic aperture radar (SAR) image of the same scene has been fused and classified. The algorithm is compared to some other advanced training algorithms in the literature for the same image data.     

Defective wafer detection using a denoising autoencoder for semiconductor manufacturing processes

Defective wafer detection is essential to avoid loss of yield due to process abnormalities in semiconductor manufacturing. For most complex processes in semiconductor manufacturing, various sensors are installed on equipment to capture process information and equipment conditions, including pressure, gas flow, temperature, and power. Because defective wafers are rare in current practice, supervised learning methods usually perform poorly as there are not enough defective wafers for fault detection (FD). The existing methods of anomaly detection often rely on linear excursion detection, such as principal component analysis (PCA), k-nearest neighbor (kNN) classifier, or manual inspection of equipment sensor data. However, conventional methods of observing equipment sensor readings directly often cannot identify the critical features or statistics for detection of defective wafers. To bridge the gap between research-based knowledge and semiconductor practice, this paper proposes an anomaly detection method that uses a denoise autoencoder (DAE) to learn a main representation of normal wafers from equipment sensor readings and serve as the one-class classification model. Typically, the maximum reconstruction error (MaxRE) is used as a threshold to differentiate between normal and defective wafers. However, the threshold by MaxRE usually yields a high false positive rate of normal wafers due to the outliers in an imbalanced data set. To resolve this difficulty, the Hampel identifier, a robust method of outlier detection, is adopted to determine a new threshold for detecting defective wafers, called MaxRE without outlier (MaxREwoo). The proposed method is illustrated using an empirical study based on the real data of a wafer fabrication. Based on the experimental results, the proposed DAE shows great promise as a viable solution for on-line FD in semiconductor manufacturing.     

基于深度学习模型的无线传感器网络数据融合算法

为了在无线传感器网络中提高数据融合性能,基于深度学习模型,提出一种将层叠自动编码器(SAE)和分簇协议相结合的数据融合算法SAEMDA,该算法在各个簇内构建特征提取分类模型SAEM,通过SAEM对节点数据进行特征提取和分类,之后将同类特征融合并发送给汇聚节点。SAEM的训练既可以采用离线有监督学习也可以采用在线无监督学习。仿真实验表明:和BPFDA,SOFMDA算法相比,SAEMDA在网络能耗大致相当的情况下能将数据融合正确率提高最多7.5%。     

一种鲁棒的半监督联邦学习系统

联邦学习允许边缘设备或客户端将数据存储在本地来合作训练共享的全局模型。主流联邦学习系统通常基于客户端本地数据有标签这一假设,然而客户端数据一般没有真实标签,且数据可用性和数据异构性是联邦学习系统面临的主要挑战。针对客户端本地数据无标签的场景,设计一种鲁棒的半监督联邦学习系统。利用FedMix方法分析全局模型迭代之间的隐式关系,将在标签数据和无标签数据上学习到的监督模型和无监督模型进行分离学习。采用FedLoss聚合方法缓解客户端之间数据的非独立同分布（non-IID）对全局模型收敛速度和稳定性的影响,根据客户端模型损失函数值动态调整局部模型在全局模型中所占的权重。在CIFAR-10数据集上的实验结果表明,该系统的分类准确率相比于主流联邦学习系统约提升了3个百分点,并且对不同non-IID水平的客户端数据更具鲁棒性。     

Content based audio classification: a neural network approach

Content based music genre classification is a key component for next generation multimedia search agents. This paper introduces an audio classification technique based on audio content analysis. Artificial Neural Networks (ANNs), specifically multi-layered perceptrons (MLPs) are implemented to perform the classification task. Windowed audio files of finite length are analyzed to generate multiple feature sets which are used as input vectors to a parallel neural architecture that performs the classification. This paper examines a combination of linear predictive coding (LPC), mel frequency cepstrum coefficients (MFCCs), Haar Wavelet, Daubechies Wavelet and Symlet coefficients as feature sets for the proposed audio classifier. Parallel to MLP, a Gaussian radial basis function (GRBF) based ANN is also implemented and analyzed. The obtained prediction accuracy of 87.3% in determining the audio genres claims the efficiency of the proposed architecture. The ANN prediction values are processed by a rule based inference engine (IE) that presents the final decision.     

面向社会事件的半监督自训练多方立场分析

已有的立场分析方法主要采用有监督或无监督方式训练立场分类模型,有监督模型训练通常需要大量有标注数据支持,而相比有监督模型,无监督模型的性能差距较大.为了降低模型训练对有标注训练数据的要求,同时保证模型性能,文中面向社会事件相关的社交媒体文本,提出半监督自训练多方立场分析方法.对于自训练方法,在模型迭代训练过程中,选择高质量样本加入训练集合,对提升模型性能起到关键作用.为此,文中方法首先根据用户立场一致性度量文本的分类置信度,然后利用话题信息进一步筛选高质量样本扩充训练集合,保证模型性能不断提升.实验表明,相比相关工作中的代表性方法和其它半监督模型训练方式,文中方法能够取得更优的立场分类效果,并且方法依据的用户立场一致性和话题信息均有助于提升立场分类效果.     

基于模式识别的传感器故障诊断

为满足模式识别故障诊断算法的鲁棒性要求,在小波包分解提取特征向量的基础上,提出了有监督模式分类与无监督模式分类相结合的故障诊断方法．利用小波包分解提取各个频带的能量作为特征向量;采用LVQ神经网络作为有监督的模式分类器进行故障诊断;运用无监督的减法聚类方法对新型故障模式进行辨识．最后,通过动力系统管路流量传感器数据对算法进行检验,验证了所提出方法的实用性和有效性．     

基于支持向量机的汉语歧义切分算法

针对于解决交集型伪歧义字段的切分,提出了一种应用支持向量机的汉语歧义切分方法.歧义切分问题可看为一个模式分类问题,为提高字段处理能力,应用支持向量机方法建立分类模型.先对歧义字段进行特征提取,采用互信息来表示歧义字段.求解过程是一个有教师学习过程,从歧义字段中挑选出一些高频伪歧义字段,人工将其正确切分作为训练样本并代入SVM训练得到一个分类模型.在分类阶段将SVM和KNN相结合构造一个新的分类器,对于待识别歧义字段代入分类器即可得到切分结果.实验证明不仅具有一定的识别准确率,而且可以提高歧义切分速度.     

基于半监督学习的数据流集成分类算法

已有的数据流分类算法多采用有监督学习,需要使用大量已标记数据训练分类器,而获取已标记数据的成本很高,算法缺乏实用性。针对此问题,文中提出基于半监督学习的集成分类算法SEClass,能利用少量已标记数据和大量未标记数据,训练和更新集成分类器,并使用多数投票方式对测试数据进行分类。实验结果表明,使用同样数量的已标记训练数据,SEClass算法与最新的有监督集成分类算法相比,其准确率平均高5。33%。且运算时间随属性维度和类标签数量的增加呈线性增长,能够适用于高维、高速数据流分类问题。     

Finger vein segmentation in infrared images using supervised and unsupervised clustering algorithms

In this paper, two new methods to segment infrared images of finger in order to perform the finger vein pattern extraction task are presented. In the first, the widespread known and used K nearest neighbor (KNN) classifier, which is a very effective supervised method for clustering data sets, is used. In the second, a novel clustering algorithm named nearest neighbor clustering algorithm (NNCA), which is unsupervised and has been recently proposed for retinal vessel segmentation, is used. As feature vectors for the classification process in both cases two features are used: the multidirectional response of a matched filter and the minimum eigenvalue of the Hessian matrix. The response of the multidirectional filter is essential for robust classification because offers a distinction between vein-like and edge-like structures while Hessian based approaches cannot offer this. The two algorithms, as the experimental results show, perform well with the NNCA has the advantage that is unsupervised and thus can be used for full automatic finger vein pattern extraction. It is also worth to note that the proposed vector, composed only of two features, is the simplest feature set which has proposed in the literature until now and results in a performance comparable with others that use a vector with much larger size (31 features). NNCA evaluated also quantitatively on a database which contains artificial images of finger and achieved the segmentation rates: 0.88 sensitivity, 0.80 specificity and 0.82 accuracy.     

Identification of non-functional requirements in textual specifications: A semi-supervised learning approach

ContextEarly detection of non-functional requirements (NFRs) is crucial in the evaluation of architectural alternatives starting from initial design decisions. The application of supervised text categorization strategies for requirements expressed in natural language has been proposed in several works as a method to help analysts in the detection and classification of NFRs concerning different aspects of software. However, a significant number of pre-categorized requirements are needed to train supervised text classifiers, which implies that analysts have to manually assign categories to numerous requirements before being able of accurately classifying the remaining ones.ObjectiveWe propose a semi-supervised text categorization approach for the automatic identification and classification of non-functional requirements. Therefore, a small number of requirements, possibly identified by the requirement team during the elicitation process, enable learning an initial classifier for NFRs, which could successively identify the type of further requirements in an iterative process. The goal of the approach is the integration into a recommender system to assist requirement analysts and software designers in the architectural design process.MethodDetection and classification of NFRs is performed using semi-supervised learning techniques. Classification is based on a reduced number of categorized requirements by taking advantage of the knowledge provided by uncategorized ones, as well as certain properties of text. The learning method also exploits feedback from users to enhance classification performance.ResultsThe semi-supervised approach resulted in accuracy rates above 70%, considerably higher than the results obtained with supervised methods using standard collections of documents.ConclusionEmpirical evidence showed that semi-supervision requires less human effort in labeling requirements than fully supervised methods, and can be further improved based on feedback provided by analysts. Our approach outperforms previous supervised classification proposals and can be further enhanced by exploiting feedback provided by analysts.     

基于激光测距AdaBoost分类的室内环境识别

移动机器人所处环境的地点语义信息能够提高机器人自主定位、路径规划和人机互动的能力.为了让机器人识别环境中不同地点类型,提出一种对机器人所处环境地点类型进行语义分类的方法.该方法对激光传感器的测距数据进行特征提取,通过提取的样本集利用强化学习AdaBoost方法构建分类器,对于环境中多类型地点分类识别,将获得的二分类器有顺序地排列建立分类列表形成多分类器,将获得的多分类器运用到房间、走廊和门口的分类识别中.实验结果表明:移动机器人通过该方法都能对环境下不同地点类型进行有效的分类识别.     

权值优化集成卷积神经网络及其在三维模型识别中的应用

三维模型应用广泛,如何有效地管理和分类这些数据库中的三维模型一直是人们 关注的问题。然而,由于不同三维模型之间的相似性难以测量,因而很难获得一种稳健且广泛 适用的三维模型分类算法。为此,提出了一种权值优化集成卷积神经网络(WOTCNN)模型,并 将其应用到三维模型的分类识别中。首先,获取三维模型的深度投影视图来最大限度地保留三维 模型的空间信息。然后,采用调整的 VGG 网络对各角度的深度投影图像进行训练并提取预测概 率值。最后,通过加权集成算法获得完整三维模型的最终分类结果。对 ModelNet10 及 ModelNet40 数据库的实验表明：三维模型的平均分类准确率达到 92.84%和 86.51%。在预测性能方面,该网 络优于普通的单卷积神经网络;在三维模型识别方面,其分类准确率能够得到显著提升。     

Ensemble of Efficient Minimal Learning Machines for Classification and Regression

Minimal Learning Machine (MLM) is a recently proposed supervised learning algorithm with performance comparable to most state-of-the-art machine learning methods. In this work, we propose ensemble methods for classification and regression using MLMs. The goal of ensemble strategies is to produce more robust and accurate models when compared to a single classifier or regression model. Despite its successful application, MLM employs a computationally intensive optimization problem as part of its test procedure (out-of-sample data estimation). This becomes even more noticeable in the context of ensemble learning, where multiple models are used. Aiming to provide fast alternatives to the standard MLM, we also propose the Nearest Neighbor Minimal Learning Machine and the Cubic Equation Minimal Learning Machine to cope with classification and single-output regression problems, respectively. The experimental assessment conducted on real-world datasets reports that ensemble of fast MLMs perform comparably or superiorly to reference machine learning algorithms.     

Pulmonary nodule diagnosis using dual‐modal supervised autoencoder based on extreme learning machine

In recent years, deep learning techniques have been applied to the diagnosis of pulmonary nodules. In order to improve the pulmonary nodule diagnostic performance effectively, we propose a novel pulmonary nodule diagnosis method using dual‐modal deep supervised autoencoder based on extreme learning machine for which discriminative features are automatically learnt from the input data. The network is fed with nodule images in pairs obtained from computed tomography and positron emission tomography respectively. For each pair image, the high‐level discriminative features of nodules in computed tomography and positron emission tomography are extracted from stacked supervised autoencoder layers. The outputs of the proposed architecture are combined using an ideal fusion method to get the final classification. In the experiments, 5‐fold cross‐validation method is used to validate the proposed method on 1,600 pulmonary nodule images and our method reaches high‐classification sensitivities of 91.75% at 1.58 false positives per scan. Meanwhile, compared with other deep learning diagnosis methods, our method achieves better discriminative results and is highly suited to be used for pulmonary nodule diagnosis.     

面向SAR目标识别的深度卷积神经网络结构设计

目的 针对用于SAR （synthetic aperture radar） 目标识别的深度卷积神经网络模型结构的优化设计难题,在分析卷积核宽度对分类性能影响基础上,设计了一种适用于SAR目标识别的深度卷积神经网络结构。方法 首先基于二维随机卷积特征和具有单个隐层的神经网络模型-超限学习机分析了卷积核宽度对SAR图像目标分类性能的影响;然后,基于上述分析结果,在实现空间特征提取的卷积层中采用多个具有不同宽度的卷积核提取目标的多尺度局部特征,设计了一种适用于SAR图像目标识别的深度模型结构;最后,在对MSTAR （moving and stationary target acquisition and recognition） 数据集中的训练样本进行样本扩充基础上,设定了深度模型训练的超参数,进行了深度模型参数训练与分类性能验证。结果 实验结果表明,对于具有较强相干斑噪声的SAR图像而言,采用宽度更大的卷积核能够提取目标的局部特征,提出的模型因能从输入图像提取目标的多尺度局部特征,对于10类目标的分类结果（包含非变形目标和变形目标两种情况）接近或优于已知文献的最优分类结果,目标总体分类精度分别达到了98.39%和97.69%,验证了提出模型结构的有效性。结论 对于SAR图像目标识别,由于与可见光图像具有不同的成像机理,应采用更大的卷积核来提取目标的空间特征用于分类,通过对深度模型进行优化设计能够提高SAR图像目标识别的精度。