MKL-SING: A data-driven approach of sign recognition for managing and improving public services

The detection and identification of traffic signs is a fundamental function of an intelligent transportation system. The extraction or identification of a road sign poses the same problems as object identification in natural contexts: conditions of illumination are variable and uncontrollable, and various objects frequently surround road signs. These difficulties make the extraction of features difficult. The fusion of time and space features of traffic signs is important for improving the performance of sign recognition. Deep learning-based algorithms are time-consuming to train based on a large amount of data. They are difficult to deploy on resource-constrained portable devices and conduct sign detection in real time. The accuracy of sign detection should be further improved, which is related to the safety of traffic participants. To improve the accuracy of feature extraction and classification of traffic signs, we propose MKL-SING, a hybrid approach based on multi-kernel support vector machine (MKL-SVM) for public transportation SIGN recognition. It contains three main components: a principal component analysis for image dimension reduction, a fused feature extractor, and a multi-kernel SVM-based classifier. The fused feature extractor extracts and fuses the time and space features of traffic signs. The multi-kernel SVM then classifies the traffic signs based on the fused features. Different kernel functions in the multi-kernel SVM are fused based on a feature weighting procedure. Compared with single-core SVM, multi-kernel SVM can better process massive data because it can project each kernel function into high-dimensional feature space to get global solutions. Finally, the performance of SVM-TSR is validated based on three traffic sign datasets. Experiment results show that SVM-TSR performs better than state-of-the-art methods in terms of dynamic traffic sign identification and recognition.     

Estimation fusion of nonlinear cost functions with application to multisensory Kalman filtering

This paper focuses on four fusion algorithms for the estimation of nonlinear cost function (NCF) in a multisensory environment. In multisensory filtering and control problems, NCF represents a nonlinear multivariate functional of state variables, which can indicate useful information of the target systems for automatic control. To estimate the NCF using multisensory information, we propose one centralized and three decentralized estimation fusion algorithms. For multivariate polynomial NCFs, we propose a simple closed-form computation procedure. For general NCFs, the most popular procedure for the evaluation of their estimates is based on the unscented transformation. The effectiveness and estimation accuracy of the proposed fusion algorithms are demonstrated with theoretical and numerical examples.     

一种新的土地覆盖类别面积估计方法及其在最大似然分类法中的应用

首先根据生产者精度和使用者精度的概念，提出生产者精度和使用者精度的条件概率表达式，然后根据概率乘积公式，推导出生产者精度和使用者精度之间的关系式。该关系式表明：①使用者精度和生产者精度的比值可作为类别真实面积与分类结果中类别面积的比值的估计;②利用使用者精度和生产者精度的比值可对遥感分类结果进行修正，产生更接近于真实值的土地覆盖类别面积值，且该方法的计算结果仅取决于使用者精度和生产者精度数据的可靠性，与分类算法的优劣无关。该方法可用于最大似然分类方法中先验概率的估计。对Erdas Imagine软件所附带lanierimg文件的实验结果表明，各种分类结果包括一种对常规最大似然分类结果进行任意修改后的分类结果，利用文中提出方法修正后均产生了接近于真实值的类别面积比例。由于作为标准的精度检验方法，几乎所有的分类影像都会产生误差矩阵用于精度报告，这保证了该方法具有很好的应用价值，可以帮助土地利用/土地覆盖研究中获取更准确的土地利用/土地覆盖的面积数据。     

Machine learning-based human-robot interaction in ITS

In the last few years, intelligent transport systems (ITS) have drawn growing attention, and these applications would have a clear and more comfortable experience for transportation. ITS provides applications with a chance to address the future condition on the route beforehand. The major issues in ITS to accomplish a precise and effective traffic flow prediction system are essential. Therefore, in this paper, a machine learning-assisted intelligent traffic monitoring system (ML-ITMS) has proposed improving transportation protection and reliability to tackle several challenges. The suggested ML-ITMS uses mathematical models to improve the accuracy estimation of traffic flow and nonparametric processes. The Machine Learning-based (ML) method is one of the best-known methods of nonparametric. It requires less prior information about connections between various traffic patterns, minor estimation limitations, and better suitability of nonlinear traffic data features. Human-Robot Interaction (HRI) helps resolve crucial issues concurrently on both the customers and service supplier levels at both ends of the transport system. Thus the experimental results show the proposed ML-ITMS to enhance traffic monitoring to 98.6% and better traffic flow prediction systems than other existing methods.     

Traffic modeling for communications networks: A multifractal approach based on few parameters

In this work, in order to model network traffic processes, an efficient multifractal model is proposed based on fewer parameters than other models present in the literature, called 2PEMV. The 2PEMV model (2-Parameter Exponential Modeling for Multipliers’ Variance) is based on a multiplicative cascade in the wavelet domain capable of synthesizing communication network traffic traces which present characteristics such as self-similarity and wide Multifractal Spectrum Width (MSW). For such a purpose, in the 2PEMV model, the energy decay of the wavelet coefficients is captured by means of an exponential modeling for the multipliers’ variance along the cascade scales. The performance of the 2PEMV model to represent the network traffic characteristics is evaluated in comparison to other models present in the literature through simulations that are carried out using real communication network traffic traces.     

Distributed state estimation and fault diagnosis using reduced sensitivity to neighbor estimates with application to building control

This paper proposes solutions that reduce the inaccuracy of distributed state estimation and consequent performance deterioration of distributed model predictive control caused by faults and inaccurate models. A distributed state estimation method for large-scale systems is introduced. A local state estimation approach considers the uncertainty of neighbor estimates, which can improve the state estimation accuracy, whereas it keeps a low network communication burden. The method also incorporates the uncertainty of model parameters which improves the performance when using simplified models. The proposed method is extended with multiple models and estimates the probability of nominal and fault behavior models, which creates a distributed fault detection and diagnosis method. An example with application to the building heating control demonstrates that the proposed algorithm provides accurate state estimates to a controller and detects local or global faults while using simplified models.     

高铁工程全生命成本智能预测模型与实证研究

为了提高高铁工程成本估算的准确性，简化高铁工程成本估算程序，本文提出了基于显著性成本理论的高铁工程全生命成本智能预测模型。首先，结合已完同类工程量清单，运用显著性成本理论，确定同类工程显著性成本项目与显著性因子；其次，运用余弦相似度法，结合已完工程与拟建工程的工程特征数据，完成相似工程的筛选；然后，结合相似工程数量的多少，分别选用基于PSO-RBF、PSO-LSSVM、FCM聚类、FIS的模糊推理预测模型进行高铁工程全生命成本预测；最后，通过实例验证该模型的准确性与可行性。研究结果表明该模型的运用确保了估算的精度，并提高了高铁工程成本估算效率。     

Joint estimation of position and attitude for intelligent vehicles by fusing delayed visual measurements

Accurate position and attitude information is an important basis for normal driving of intelligent vehicles. In this paper, we investigate the estimation of position and attitude states for intelligent vehicles with low cost scheme. The low cost GNSS, camera, and proprioceptive sensors equipped by mass-produced vehicle are fused to estimate the states. The visual measurements adopted in this paper are based on the lateral distance and deflection angle to road features such as lane lines or curbs, which are generated more frequently than some other semantic features such as traffic lights, and leads to broader application scenario. Moreover, it is easier to implement compared with geometrical feature matching methods, since it only needs a simple prior map while latter needs large maps containing many high precision features. The visual measurements is often with large time delay due to negligible processing time. In order to fuse delayed measurements, a state-augmentation technique is adopted for the estimator design. The performance of the proposed method is evaluated by professional simulation software CarMaker, and shows that the incorporation of road features based visual measurement can effectively improve the position and attitude estimation accuracy by reducing the lateral position and yaw angle estimation error.     

基于核密度相关度量的视频目标跟踪

传统的Mean shift 方法采用颜色直方图作为特征，以Bhattacharyya系数作为目标参考模板与当前帧中候选目标间的相似度量，通过迭代寻找距离函数的局部最小值，从而得到当前帧中的目标实际位置。由于颜色直方图仅仅描述了图像中目标的全局颜色分布而忽略了空间位置分布，使得当目标邻域中存在与目标相近似的颜色模式时，算法无法取得理想的跟踪效果。本文提出了基于核密度估计相关的距离度量，在描述参考目标和候选目标时，考虑到诸如颜色、梯度等目标像点的特征区间的同时，融入了目标像点的空间位置信息，使得跟踪算法更加稳健和精确，能够更好适应目标和背景变化复杂的应用场合。     

Bayesian and frequentist estimation of the performance of free space optical channels under weak turbulence conditions

Approximate mathematical form expressions were derived for the estimation of the average (ergodic) capacity and the average bit error rate of a log-normal free space optical channel in the cases of weak to moderate atmospheric turbulence conditions. We investigate the average capacity, the average bit error rate and the outage probability of free space optical communication channels using the frequentist and the Bayesian approach. Emphasis is given on the cases of weak to moderate atmospheric turbulence leading to channels modeled by log-normal distributed intensity fading. Furthermore, accurate approximate closed-form expressions and estimation procedures for their achievable capacity as well as their bit error rate and the important parameters of interest are derived. The derived approximate analytical expressions are verified by various numerical examples and simulations. Moreover, each methodology is reviewed in terms of their analytic convenience and their accuracy is also discussed.     

Target layer state estimation in multilayer complex dynamical networks using functional observability

In large-scale complex dynamical networks, it is significant to estimate the states of target nodes with only a part of measured nodes. Meanwhile, multilayer complex dynamical networks exist widely in society and engineering. Therefore, it has important theoretic meaning and practical value to study the state estimation of target nodes in multilayer complex dynamical networks with limited node measurements. In this paper, with the measurable state information of a portion of nodes in one layer in the multilayer complex dynamical network, the state estimation of target nodes in other layers is studied. First, we build the model of the multilayer complex dynamical network which includes some target nodes and sensor nodes. Second, auxiliary nodes are selected by using the maximum matching principle in graph theory to construct the augmented node set. Third, we discuss the relationship between the minimum number of auxiliary nodes and interlayer connection probability in the multilayer complex dynamical network. Forth, an appropriate functional state observer is designed with limited number of measured nodes according to a typical model-based algorithm. Finally, numerical simulations are given to demonstrate the accuracy of the proposed method. The proposed method can achieve the accurate estimation with less placement of observers and fewer computational costs in the multilayer complex dynamical network.     

嵌入式视频车辆检测系统

设计并实现了一种基于DM642视频车辆检测系统。提出了一种基于虚拟线圈匹配的算法,算法原理是估计车辆的速度和结合背景差统计量来检测通过监测区域的车辆,把算法移植到DM642平台中进行处理运行,处理结果可以满足视频处理的实时性要求,并且算法对外界环境变化的干扰有一定的鲁棒性,准确的车流量检测精度,并且运行稳定。     

Performance analysis of a spreading sequence estimator for spread spectrum transmissions

Direct sequence spread spectrum transmissions (DS-SS) are now widely used for secure communications, as well as for multiple access. They have many interesting properties, including low probability of interception. Indeed, DS-SS transmitters use a periodical pseudo-random sequence to modulate the baseband signal before transmission. A receiver which does not know the sequence cannot demodulate the signal.In this paper, we propose a new method which can estimate the spreading sequence in a noncooperative context. The method is based on eigenanalysis techniques. The received signal is divided into windows, from which a covariance matrix is computed. We show that the sequence can be reconstructed from the two first eigenvectors of this matrix, and that useful information, such as desynchronisation time, can be extracted from the eigenvalues.The main achievement of the present paper is a performance analysis of the proposed spreading sequence estimation procedure. An analytical approach is first considered owing to matrix perturbation theory and Wishart matrix properties. Then, complementary Monte Carlo simulations are performed to show the effectiveness of the proposed method.     

DNS tunnels detection via DNS-images

DNS tunneling is a typical attack adopted by cyber-criminals to compromise victims’ devices, steal sensitive data, or perform fraudulent actions against third parties without their knowledge. The fraudulent traffic is encapsulated into DNS queries to evade intrusion detection. Unfortunately, traditional defense systems based on Deep Packet Inspection cannot always detect such traffic. As a result, DNS tunneling is one problem that has worried the cybersecurity community over the past decade.In this paper, we propose a robust and reliable Deep Learning-based DNS tunneling detection approach to mine valuable insight from DNS query payloads. More precisely, several features are first extracted by the DNS flow, and then they are arranged as bi-dimensional images. A Convolutional Neural Network is used to automatically and adaptively learn spatial hierarchies of features to be used in a fully connected neural network for traffic classification. The proposed approach may result in an extremely interesting task in predictive security approaches to attack detection.The effectiveness of the proposal is evaluated in several experiments using a real-world traffic dataset. The obtained results show that our approach achieves 99.99% of accuracy and performs better than state-of-the-art solutions.     

交通数据中频繁轨迹模式挖掘

在当今类似车载系统中常有需要我们在物体时空运动数据库中进行数据挖掘,然后根据得到的规则预测物体的运动趋势的需求。至今,对于由时空两方面描述的数据进行挖掘的研究还没有特别明确的方法。提出MINE_ALLFP算法来发现所有的频繁集。为了增加研究的可行性,去掉位置信息的连续性,提出将整个大空间分割为若干个小区域的方法,这里还引进一种好的剪枝算法。     

Connecting user and item perspectives in popularity debiasing for collaborative recommendation

Recommender systems learn from historical users’ feedback that is often non-uniformly distributed across items. As a consequence, these systems may end up suggesting popular items more than niche items progressively, even when the latter would be of interest for users. This can hamper several core qualities of the recommended lists (e.g., novelty, coverage, diversity), impacting on the future success of the underlying platform itself. In this paper, we formalize two novel metrics that quantify how much a recommender system equally treats items along the popularity tail. The first one encourages equal probability of being recommended across items, while the second one encourages true positive rates for items to be equal. We characterize the recommendations of representative algorithms by means of the proposed metrics, and we show that the item probability of being recommended and the item true positive rate are biased against the item popularity. To promote a more equal treatment of items along the popularity tail, we propose an in-processing approach aimed at minimizing the biased correlation between user-item relevance and item popularity. Extensive experiments show that, with small losses in accuracy, our popularity-mitigation approach leads to important gains in beyond-accuracy recommendation quality.     

概率估算方法对中文文本特征提取质量的影响

在文本自动分类中,目前有词频和文档频率统计这两种概率估算方法,采用的估算方法恰当与否会直接影响特征抽取的质量与分类的准确度。本文采用K最近邻算法实现中文文本分类器,在中文平衡与非平衡两种训练语料下进行了训练与分类实验,实验数据表明使用非平衡语料语料时,可以采用基于词频的概率估算方法,使用平衡语料语料时,采用基于文档频率的概率估算方法,能够有效地提取高质量的文本特征,从而提高分类的准确度。     

Probability-constrained tracking control for a class of time-varying nonlinear stochastic systems

This paper is concerned with the probability-constrained tracking control problem for a class of time-varying systems with stochastic nonlinearities, stochastic noises and successively packet loss. The main purpose of this paper is to design a time-varying observer and tracking controller such that (1) the probabilities of both the estimation error and tracking error confined to given ellipsoidal sets are larger than prescribed constants, and (2) the ellipsoids are minimized in the sense of matrix norm at each time point. By using a stochastic analysis method, the probability constrained tracking control problem is solved and sufficient conditions are obtained in terms of recursive linear matrix inequalities. A recursive optimization algorithm is developed to design the observer and tracking controller such that not only the addressed probability constrained aim is satisfied, but also the ellipsoidal sets are minimized. At last, a simulation example is given to illustrate the effectiveness and applicability of the developed approach.     

散粒体自组织临界性的多重分形分析

简述了自组织临界性理论的基本概念和多重分形原理方法,介绍了大尺度的非均匀沙堆模型实验,并表明其呈现出自组织临界性特征。并对沙堆系统的时空分形结构进行分析计算．确定了广义维数和奇异指数,得出多重分形特征谱。为研究沙堆的演化提出了多重分形动力模型。最后阐明了采用多重分形方法进行散粒体自组织临界性研究的意义。     

Extracting translations from comparable corpora for Cross-Language Information Retrieval using the language modeling framework

A main challenge in Cross-Language Information Retrieval (CLIR) is to estimate a proper translation model from available translation resources, since translation quality directly affects the retrieval performance. Among different translation resources, we focus on obtaining translation models from comparable corpora, because they provide appropriate translations for both languages and domains with limited linguistic resources. In this paper, we employ a two-step approach to build an effective translation model from comparable corpora, without requiring any additional linguistic resources, for the CLIR task. In the first step, translations are extracted by deriving correlations between source–target word pairs. These correlations are used to estimate word translation probabilities in the second step. We propose a language modeling approach for the first step, where modeling based on probability distribution provides two key advantages. First, our approach can be tuned easier in comparison with heuristically adjusted previous work. Second, it provides a principled basis for integrating additional lexical and translational relations to improve the accuracy of translations from comparable corpora. As an indication, we integrate monolingual relations of word co-occurrences into the process of translation extraction, which helps to extract more reliable translations for low-frequency words in a comparable corpus. Experimental results on an English–Persian comparable corpus show that our method outperforms the previous approaches in terms of both translation quality and the performance of CLIR. Indeed, the proposed method is naturally applicable to any comparable corpus, regardless of its languages. In addition, we demonstrate the significant impact of word translation probabilities, estimated in the second step of our approach, on the performance of CLIR.