CBS: Contourlet-Based Steganalysis Method

An ideal steganographic technique embeds secret information into a carrier cover object with virtually imperceptible modification of the cover object. Steganalysis is a technique to discover the presence of hidden embedded information in a given object. Each steganalysis method is composed of feature extraction and feature classification components. Using features that are more sensitive to information hiding yields higher success in steganalysis. So far, several steganalysis methods have been presented which extract some features from DCT or wavelet coefficients of images. Multi-scale and time-frequency localization of an image is offered by wavelets. However, wavelets are not effective in representing the images in different directions. Contourlet transform addresses this problem by providing two additional properties, directionality and anisotropy. The present paper offers an universal approach to steganalysis called CBS, which uses statistical moments of contourlet coefficients as features for analysis. After feature extraction, a non-linear SVM classifier is applied to classify cover and stego images. The efficiency of the proposed method is demonstrated by experimental investigations. The proposed steganalysis method is compared with two well-known steganalyzers against typical steganography methods. The results showed the superior performance of our method.     

A blind audio steganalysis based on feature fusion

In this paper, we present a blind steganalysis based on feature fusion. Features based on Short Time Fourier Transform (STFT), which consists of second-order derivative spectrum features of audio and Mel-frequency cepstrum coefficients, audio quality metrics and features on linear prediction residue are extracted separately. Then feature fusion is conducted. The performance of the proposed steganalysis is evaluated against 4 steganographic schemes: Direct Sequence Spread Spectrum (DSSS), Quantization Index Modulation (QIM), ECHO embedding (ECHO), and Least Significant Bit embedding (LSB). Experiment results show that the classifying performance of the proposed detector is much superior to the previous work. Even more exciting is that the proposed methodology could detect the four steganography, with 85%+ classification accuracy achieved in all the detections, which makes the proposed steganalysis methodology capable of being regarded as a blind steganalysis, and especially useful when the steganalyzer are without the knowledge of the steganographic scheme employed in data embedding.     

Mixing high-dimensional features for JPEG steganalysis with ensemble classifier

This paper proposes a JPEG steganalysis scheme based on the ensemble classifier and high-dimensional feature space. We first combine three current feature sets and remove the unimportant features according to the correlation between different features parts so as to form a new feature space used for steganalysis. This way, the dependencies among cover and steganographic images can be still represented by the features with a reduced dimensionality. Furthermore, we design a proportion mechanism to manage the feature selection in two subspaces for each base learner of the ensemble classifier. Experimental results show that the proposed scheme can effectively defeat the MB and nsF5 steganographic methods and its performance is better than that of existing steganalysis approaches.     

一种基于DCT域统计特征的JPEG图像隐写分析

提出了一种基于DCT域统计特征的JPEG图像隐写分析算法。该算法在分析JPEG图像的DCT域统计特性的基础上,提取了8维特征向量,通过LSSVM分类器对待测图像进行分类,以达到检测载密图像的目的。算法实现简单、计算复杂度低。实验结果表明,该算法检测速度快,具有较高的检测正确率,能够实现针对各类JPEG图像信息隐藏算法的有效检测。     

一种动态赋权红外光谱特征选择算法研究

大规模的红外光谱数据集中存在大量无关冗余的特征。针对这一问题,提出了一种动态赋权红外光谱特征选择算法(Dynamic Weight Infrared Spectrum Feature Selection Algorithm, MBDWFS)。 该算法把对称不确定性度量标准与近似Markov Blanket相结合,以删除原始光谱数据集中无关冗余的特征,从而获取数据规模较小且最优的特征子集。通过与 FCBF、ID$_3$ 和ReliefF三种经典特征选择算法的性能仿真对比试验,证明所提出的MBDWFS算法在整体分类性能上优于其他三种算法,用于红外光谱的物质分析领域时效果更好。     

Feature selection algorithm based on XGBoost

Feature selection in classification has always been an important but difficult problem.This kind of problem requires that feature selection algorithms can not only help classifiers to improve the classification accuracy,but also reduce the redundant features as much as possible.Therefore,in order to solve feature selection in the classification problems better,a new wrapped feature selection algorithm XGBSFS was proposed.The thought process of building trees in XGBoost was used for reference,and the importance of features from three importance metrics was measured to avoid the limitation of single importance metric.Then the improved sequential floating forward selection (ISFFS) was applied to search the feature subset so that it had high quality.Compared with the experimental results of eight datasets in UCI,the proposed algorithm has good performance.     

基于改进增强特征选择算法的特征融合图像隐写分析

针对现有的基于特征融合的JPEG隐写分析方法特征冗余度高、通用性较低的问题,提出了一种基于改进的增强特征选择(BFS,boosting feature selection)算法的通用JPEG隐写分析方法。从线性相关度和非线性相关度两方面降低特征冗余,将特征自相关系数和互信息这两种统计性能引入到特征的评价准则中,重新设计了特征权重计算方法,改进了BFS算法的特征评价函数。通过改进的BFS特征选择算法将3组互补性较强且准确率高的特征进行融合降维,得到最优特征子集训练分类器。对3种高隐蔽性隐写算法F5、Outguess和MME3,在不同嵌入率下进行了大量实验。结果表明,本文方法的分析准确率高于现有的检测率较高的JPEG隐写分析方法和典型的融合分析方法,融合后的特征相关性明显下降,并且具有更强的通用性。     

基于非平衡Fisher判别的两类红外特征提取

红外目标特征提取和识别中,由于小样本数据选择特征会存在严重的非平衡性,降低特征识别性能.基于此,提出了一种改进型非平衡Fisher判别的两类特征提取方法,以特征统计方差作为判定标准,选择两类目标特征,并对所选特征进行加权判别,以降低小样本库中目标与非目标不平衡性对识别的影响.通过一组空中目标的红外序列验证,说明了改进的加权非平衡Fisher判别标准对空中目标特征提取的有效性.     

A generalized machine learning-based model for the detection of DDoS attacks

As time is progressing, the number and the complexity of methods adopted for launching distributed denial of service (DDoS) attacks are changing. Therefore, we propose a methodology for the development of a generalized machine learning (ML)-based model for the detection of DDoS attacks. After exploring various attributes of the dataset chosen for this study, we propose an integrated feature selection (IFS) method which consists of three stages and integration of two different methods, that is, filter and embedded methods to select features which highly contribute to the detection of various types of DDoS attacks. We use light gradient boosting machine (LGBM) algorithm for training of the model for classification of benign and malicious flows. For ensuring satisfactory performance and generalized behavior of the developed model, we test it by passing records of unseen DDoS attack types. Several performance metrics are employed for the evaluation of the model. By comparing the performance of developed model against state-of-the-art models, we state an improvement of around 20% for almost all the reported metrics. We also show that the performance of the model improves if feature space is reduced by 77%. Furthermore, the generalized behavior of the developed model is justified by demonstrating a trade-off between high variance and high bias ML models.     

基于LGBM的Zernike特征选取及红外图像目标识别方法

红外传感技术有效解决了夜间观测的难题,成为现代战场侦察的重要手段之一。不断提升基于红外图像的目标识别能力是实施精确打击、态势感知的有力途径。针对红外图像识别问题,提出基于轻量级梯度提升机（Light Gradient Boosting Machine, LGBM）的Zernike特征选取算法,并结合稀疏表示分类器（Sparse Representation-based Classification, SRC）完成目标类别确认。首先,基于红外图像中的目标区域提取多阶Zernike矩特征,表征待识别目标的本质特性;其次,采用LGBM特征选择算法对多阶矩特征进行二次筛选,减少冗余的同时提高特征的针对性;最后,基于SRC对最终选择的Zernike矩特征矢量进行分类。该方法通过LGBM的特征选择有效提高了最终特征的有效性,同时降低了分类的计算复杂度,有利于提高整体识别性能。采用公开的中波红外目标图像数据集（MWIR）开展验证实验,对10类典型军事目标进行区分识别。实验分别在原始样本、噪声干扰样本以及部分缺失样本三种条件下进行并与几类现有红外目标识别方法进行对比讨论。结果表明:所提方法可取得更优性能,证明其有效性。     

Interest Points of General Imbalance

The imbalance oriented selection scheme was recently introduced to detect stable interest points in weakly or sparsely textured images. The scheme chooses image points whose one-pixel-wide directional intensity variations can be clustered into two imbalanced classes as candidates. An important property of imbalance oriented selection is that imbalanced points can be contiguous to others, i.e., imbalanced points have local geometry coherent property. In this paper, we propose general imbalance decided by multipixel-wide directional intensity variations. We give a theoretical analysis on a relation between imbalance and general imbalance. In terms of the local geometry coherent property of general imbalanced points, we propose a global-to-local appearance based matching scheme for imbalanced point correspondence. Last, we present an application of general imbalanced points to road sign detection, which demonstrates the good potential of general imbalanced points.     

Unbalanced JPEG image steganalysis via multiview data match

Image steganalysis must address the matter of learning from unbalanced training sets where the cover objects (normal images) always greatly outnumber the stego ones. But the research in unbalanced image steganalysis is seldom seen. This work just focuses on the problem of unbalance JPEG images steganalysis. In this paper, we propose a frame of feature dimension reduction based semi-supervised learning for high-dimensional unbalanced JPEG image steganalysis. Our method uses standard steganalysis features, and selects the confident stego images from the unlabeled examples by multiview match resampling method to rebalance the unbalanced training images. Furthermore, weighted Fisher linear discriminant (WFLD) is proposed to find the proper feature subspace where K-means provides the weight factor for WFLD in return. Finally, WFLD and K-means both work in an iterative fashion until convergence. Experimental results on the MBs and nsF5 steganographic methods show the usefulness of the developed scheme over current popular feature spaces.     

面向类不平衡网络流量的特征选择算法

针对网络流量分类过程中出现的类不平衡问题,该文提出一种基于加权对称不确定性(WSU)和近似马尔科夫毯(AMB)的特征选择算法。首先,根据类别分布信息,定义了偏向于小类别的特征度量,使得与小类别具有强相关性的特征更容易被选择出来;其次,充分考虑特征与类别间、特征与特征之间的相关性,利用加权对称不确定性和近似马尔科夫毯删除不相关特征及冗余特征;最后,利用基于相关性度量的特征评估函数以及序列搜索算法进一步降低特征维数,确定最优特征子集。实验表明,在保证算法整体分类精确率的前提下,算法能够有效提高小类别的分类性能。     

基于分数阶傅里叶变换的窄带雷达飞机目标回波特征提取方法

该文研究了常规窄带雷达体制下,利用分数阶傅里叶变换扩展特征域,从而解决直升机、螺旋桨飞机和喷气式飞机3类飞机目标回波分类中的特征提取问题。在现代战场中,直升机、螺旋桨飞机和喷气式飞机具有不同的机动性能,并各自承担着重要的任务。因此,实现这3类飞机的分类具有重大的意义。该文针对3类飞机目标分类的传统特征数目少,包含信息量有限,导致分类性能不够好的问题,基于现有的特征提取方法引入分数阶傅里叶变换(Fractional Fourier Transform, FrFT),在经过FrFT后的分数域提取3类飞机目标回波的分数阶特征,弥补传统特征的不足。并利用线性相关向量机(Relevance Vector Machine, RVM)的特征选择功能对提取的分数阶特征进行特征选择并分类。基于仿真和实测数据的实验结果证明该文提出的分数阶特征的分类性能较传统时域、多普勒域特征有较大提升。     

面向类不平衡网络流量的特征选择算法

针对网络流量分类过程中出现的类不平衡问题,该文提出一种基于加权对称不确定性(WSU)和近似马尔科夫毯(AMB)的特征选择算法。首先,根据类别分布信息,定义了偏向于小类别的特征度量,使得与小类别具有强相关性的特征更容易被选择出来;其次,充分考虑特征与类别间、特征与特征之间的相关性,利用加权对称不确定性和近似马尔科夫毯删除不相关特征及冗余特征;最后,利用基于相关性度量的特征评估函数以及序列搜索算法进一步降低特征维数,确定最优特征子集。实验表明,在保证算法整体分类精确率的前提下,算法能够有效提高小类别的分类性能。     

Region based Image Steganalysis using Artificial Bee Colony

Steganalysis is the art and skill of discriminating stego images from cover images. Image steganalysis algorithms can be divided into two broad categories, specific and universal. In this paper, a novel universal image steganalysis algorithm is proposed which is called RISAB, Region based Image Steganalysis using Artificial Bee colony. The goal of the proposed method is to realize a sub-image from stego and cover images through ABC with respect to density according to the cover, stego and difference images. In our method, we look for the best sub-image, which contains the highest density with respect to the changed embedding pixels. Furthermore, after selecting the best sub-image, we extract the features, which have been selected by IFAB, Image steganalysis based on Feature selection using Artificial Bee colony. At the end, both selected features by IFAB and extracted features by RISAB are combined. As a result, a feature vector is generated which improves accuracy of steganalysis. Experimental results show that our proposed method outperforms other approaches.     

Feature Selection for Genomic Signal Processing: Unsupervised,Supervised, and Self-Supervised Scenarios

An effective data mining system lies in the representation of pattern vectors. For many bioinformatic applications, data are represented as vectors of extremely high dimension. This motivates the research on feature selection. In the literature, there are plenty of reports on feature selection methods. In terms of training data types, they are divided into the unsupervised and supervised categories. In terms of selection methods, they fall into filter and wrapper categories. This paper will provide a brief overview on the state-of-the-arts feature selection methods on all these categories. Sample applications of these methods for genomic signal processing will be highlighted. This paper also describes a notion of self-supervision. A special method called vector index adaptive SVM (VIA-SVM) is described for selecting features under the self-supervision scenario. Furthermore, the paper makes use of a more powerful symmetric doubly supervised formulation, for which VIA-SVM is particularly useful. Based on several subcellular localization experiments, and microarray time course experiments, the VIA-SVM algorithm when combined with some filter-type metrics appears to deliver a substantial dimension reduction (one-order of magnitude) with only little degradation on accuracy.     

Steganalysis using image quality metrics

We present techniques for steganalysis of images that have been potentially subjected to steganographic algorithms, both within the passive warden and active warden frameworks. Our hypothesis is that steganographic schemes leave statistical evidence that can be exploited for detection with the aid of image quality features and multivariate regression analysis. To this effect image quality metrics have been identified based on the analysis of variance (ANOVA) technique as feature sets to distinguish between cover-images and stego-images. The classifier between cover and stego-images is built using multivariate regression on the selected quality metrics and is trained based on an estimate of the original image. Simulation results with the chosen feature set and well-known watermarking and steganographic techniques indicate that our approach is able with reasonable accuracy to distinguish between cover and stego images.     

数据集类别不平衡性对迁移学习的影响分析

数据集类别不平衡性是机器学习领域的常见问题,对迁移学习也不例外。本文针对迁移学习下数据集类别不平衡性的影响研究不足的问题,重点研究了以下几种不平衡性处理方法对迁移学习的影响效果分析:过采样、欠采样、加权随机采样、加权交叉熵损失函数、Focal Loss函数和基于元学习的L2RW（Learning to Reweight）算法。其中,前三种方法通过随机采样消除数据集的不平衡性,加权交叉熵损失函数和Focal Loss函数通过调整传统分类算法的损失函数以适应不平衡数据集的训练,L2RW算法则采用元学习机制动态调整样本权重以实现更好的泛化能力。大量实验结果表明,在上述各种不平衡性处理方法中,过采样处理和加权随机采样处理更适合迁移学习。      

RO PUF硬件物理指纹性能评估

入网设备的身份认证是工业互联网、物联网安全运行的基础.通过分析物理不可克隆函数(Physical Unclonable Function,PUF)作为入网设备的身份凭据应具有的特性,建立了其性能的多层指标体系.按照性能指标的重要性分为必要指标、重要指标和应用指标,对每项指标及其评估方法进行了说明,并给出了可用于PUF选型的综合评估方案.实验结果表明,该PUF性能指标体系和评估方案在工程上是可行的.