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.     

Sparse representation for robust face recognition by dictionary decomposition

Sparse representation-based classification (SRC) method has gained great success in face recognition due to its encouraging and impressive performance. However, in SRC the data used to train or test are usually corrupted, and hence the performance is affected. This paper proposes a robust face recognition approach by means of learning a class-specific dictionary and a projection matrix. Firstly, the training data are decomposed into class-specific dictionary, non-class-specific dictionary, and sparse error matrix. Secondly, in order to correct the corrupted test data, the data are projected onto their corresponding underlying subspace, and a projection matrix between the original training data and the class-specific dictionary is learned. Then, the features of the class-specific dictionary and the corrected test data are extracted by using Eigenface method. Finally, the SRC is performed to classify. Extensive experiments conducted on publicly available data sets show that the proposed algorithm performs better than some state-of-the-art methods.     

New framework for unsupervised universal steganalysis via SRISP-aided outlier detection

Formulating steganalysis as a binary classification problem has been highly successful. However, the existing detection algorithms are difficult to obtain high detection accuracy when applied in real-world circumstances. Because so-called model mismatch problem often occurs owing to unknown cover source and embedding parameters. To avoid the mess of model mismatch, we propose a new unsupervised universal steganalysis framework to detect individual stego images. First, cover images with statistical properties similar to those of the given test image are searched from a retrieval cover database to establish an aided cover sample set. Second, unsupervised outlier detection is performed on a test set composed of the given test image and its aided cover sample set to determine the type (cover or stego) of the given test image. Our proposed framework, called Similarity Retrieval of Image Statistical Properties (SRISP)-aided unsupervised outlier detection, requires no training, and thus it does not suffer from model mismatch. The framework employs standard steganalysis features and detects each test image individually. Experimental results illustrate that the framework substantially outperforms one-class support vector machine and the traditional unsupervised outlier detectors without considering SRISP; its detection performance is independent of the proportion of stego images in the test samples.     

一种信源个数与波达方向联合估计的新算法

针对多级维纳滤波器(MSWF)用于子空间估计时信号特征矢量泄漏到噪声子空间的问题,提出了一种新的信号子空间估计算法,该算法不需要训练信号和信源个数的先验知识.随后,给出了一种信源个数的后判断方法,最终完成信源个数及方向的同时估计.整个算法不需要协方差矩阵的计算和特征值的分解,具有较低的计算复杂度.在均匀线阵且信号互不相关情况下,改进后的算法用于波达方向估计时拥有与基于特征分解方法近似的性能.仿真结果验证了该方法的有效性.     

基于谐波提取的欠定语音盲分离方法

现有的欠定语音信号盲分离算法往往不能同时兼顾分离性能及效率。针对此问题,本文提出一种基于谐波提取的欠定盲分离方法。首先,利用频谱校正从混合信号的短时傅立叶变换中提取谐波参数,其次利用相位一致性准则甄别这些参数的单源属性,进而用自适应K-均值方法对单源模式做聚类而获得源数估计和混合矩阵估计,最后再用子空间投影法恢复源信号。其中谐波提取和单源参数筛选可保证低复杂度地精确估计出混合矩阵。仿真实验表明,相比于原始子空间投影算法,本文方法可获得更高的信号恢复质量,且在谐波相关领域也具有潜在应用价值。      

基于流形等距映射的矩阵信息几何检测方法

雷达目标检测常面临复杂的杂波特性,经典的检测方法通常适合于某些特定的场景,当检测背景发生变化时,其检测性能急剧下降。为有效提升不同杂波背景下的检测性能,提出一种基于流形等距映射(ISOMAP)的矩阵信息几何检测器。该方法首先将信号检测问题转化为矩阵流形上两点之间的区分性问题;然后基于样本数据和流形等距映射原理,自适应地学习出矩阵流形的投影变换矩阵,将矩阵流形变换为可区分的低维流形,最大程度地保持每一个矩阵与其邻域内矩阵之间几何距离大小,增强矩阵流形的可分性;最后利用仿真杂波和实测数据对算法进行验证。实验结果表明,相比于经典的检测方法,所提方法能有效提升目标检测性能。     

部分均匀环境中训练样本不足时的贝叶斯子空间检测器

为了解决部分均匀环境中训练数据不足时的子空间信号检测难题,采用贝叶斯理论,将噪声协方差矩阵建模为逆威沙特分布,并采用广义似然比准则（generalized likelihood ratio test,GLRT）、Rao准则和Wald准则设计自适应检测器,结果表明3种准则得到相同的结果。基于仿真及实测数据验证了所提检测器的有效性,并得出了影响检测性能的关键物理量。     

LSA based multi-instance learning algorithm for image retrieval

Focusing on the problem of natural image retrieval, based on latent semantic analysis (LSA) and support vector machine (SVM), a novel multi-instance learning (MIL) algorithm is proposed, where a bag corresponds to an image and an instance corresponds to the low-level visual features of a segmented region. Firstly, in order to transform every bag into a single sample, a collection of “visual-word” is generated by k-means clustering method to construct a projection space, then a nonlinear mapping is defined using these “visual-word” to embed each bag as a point in the projection space, thereby obtaining every bag's projection feature. Secondly, the matrix consisted of all the projection features of training bags is regarded as a term-document matrix, and LSA method is used to obtain the latent semantic feature of each bag. As a result, the MIL problem is converted into a standard single instance learning (SIL) problem that can be solved directly by SVM method. Experimental results on the COREL data sets show that the proposed method, named LSASVM-MIL, is robust, and its performance is superior to other key existing MIL algorithms.     

Particle Swarm Optimization of Compression Measurement for Signal Detection

In this paper, the random matrix in the compressive and subspace compressive detectors is optimized based on the particle swarm optimization (PSO). The PSO, which belongs to swarm intelligent theory, is used for the first time to solve the optimization problem of the random projection matrix, leading to an improved version of the conventional compressive and subspace compressive detectors. Simulation results show the proposed PSO-based detectors can achieve a better detection performance and require fewer measurements than the traditional compressive detectors without using PSO.     

一种基于PSOSVM的盲隐写分析方法

针对盲隐写分析中的特征选择问题,提出了结合粒子群优化算法（PSO）的支持向量机分类器进行特征选择的方法。该方法使用非线性支持向量机作为分类器,使用PSO为支持向量机寻找最优的图像特征集合作为训练集和测试集,同时选择最优的支持向量机参数,进而利用最优的特征集和支持向量机参数对隐写图像进行检测。实验结果表明,该优化方法明显优于Farid。ANOVA和F—score方法,提高了检测隐写图像的成功率和系统检测效率。     

训练样本不足时的子空间信号检测方法

为了解决训练样本不足时的子空间信号检测问题,提出了两种有效的降秩检测器.基于主分量分析(PCA)的思想,先把常规自适应子空间检测器中采样协方差矩阵(SCM)的求逆运算用噪声特征子空间矩阵与其共轭转置的乘积代替,构造降秩子空间检测器;为进一步提高算法稳健性,把降秩子空间检测器的求逆运算用Moore-Penrose逆代替.仿真结果表明,所提方法在训练样本充足及不足时,均比现有方法具有更好的检测性能.     

Adaptive subspace detection of extended target in white Gaussian noise using sinc basis

For the high resolution radar (HRR), the problem of detecting the extended target is considered in this paper. Based on a single observation, a new two-step detection based on sparse representation (TSDSR) method is proposed to detect the extended target in the presence of Gaussian noise with unknown covariance. In the new method, the Sinc dictionary is introduced to sparsely represent the high resolution range profile (HRRP). Meanwhile, adaptive subspace pursuit (ASP) is presented to recover the HRRP embedded in the Gaussian noise and estimate the noise covariance matrix. Based on the Sinc dictionary and the estimated noise covariance matrix, one step subspace detector (OSSD) for the first-order Gaussian (FOG) model without secondary data is adopted to realise the extended target detection. Finally, the proposed TSDSR method is applied to raw HRR data. Experimental results demonstrate that HRRPs of different targets can be sparsely represented very well with the Sinc dictionary. Moreover, the new method can estimate the noise power with tiny errors and have a good detection performance.     

Least squares subspace projection approach to mixed pixelclassification for hyperspectral images

An orthogonal subspace projection (OSP) method using linear mixture modeling was recently explored in hyperspectral image classification and has shown promise in signature detection, discrimination, and classification. In this paper, the OSP is revisited and extended by three unconstrained least squares subspace projection approaches, called signature space OSP, target signature space OSP, and oblique subspace projection, where the abundances of spectral signatures are not known a priori but need to be estimated, a situation to which the OSP cannot be directly applied. The proposed three subspace projection methods can be used not only to estimate signature abundance, but also to classify a target signature at subpixel scale so as to achieve subpixel detection. As a result, they can be viewed as a posteriori OSP as opposed to OSP, which can be thought of as a priori OSP. In order to evaluate these three approaches, their associated least squares estimation errors are cast as a signal detection problem ill the framework of the Neyman-Pearson detection theory so that the effectiveness of their generated classifiers can be measured by receiver operating characteristics (ROC) analysis. All results are demonstrated by computer simulations and Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) data     

Interference mitigation in STAP using the two-dimensional Wold decomposition model

We propose a novel parametric approach for modeling, estimation, and detection in space-time adaptive processing (STAP) radar systems. The proposed parametric interference mitigation procedures can be applied even when information in only a single range gate is available, thus achieving high performance gain when the data in the different range gates cannot be assumed stationary. The model is based on the Wold-like decomposition of two-dimensional (2D) random fields. It is first shown that the same parametric model that results from the 2D Wold-like orthogonal decomposition naturally arises as the physical model in the problem of space-time processing of airborne radar data. We exploit this correspondence to derive computationally efficient fully adaptive and partially adaptive detection algorithms. Having estimated the models of the noise and interference components of the field, the estimated parameters are substituted into the parametric expression of the interference-plus-noise covariance matrix. Hence, an estimate of the fully adaptive weight vector is obtained, and a corresponding test is derived. Moreover, we prove that it is sufficient to estimate only the spectral support parameters of each interference component in order to obtain a projection matrix onto the subspace orthogonal to the interference subspace. The resulting partially adaptive detector is simple to implement, as only a very small number of unknown parameters need to be estimated, rather than the field covariance matrix. The performance of the proposed methods is illustrated using numerical examples.     

高光谱遥感图像端元提取的零空间光谱投影算法

端元提取技术是高光谱遥感图像光谱解混的关键.在线性光谱混合分析中,首先引入了高光谱遥感图像经过零空间光谱投影后具有单形体的凸不变性.在此基础上,提出了零空间光谱投影算法,通过设计各种度量和准则,制定不同的单次端元提取策略,灵活地实现算法.经过证明,零空间光谱投影算法是对基于子空间投影距离算法(包括零空间投影距离算法与经典正交子空间投影算法)的进一步延伸,提供了更多的端元提取策略.实验结果表明,零空间光谱投影算法在模拟图像以及真实高光谱遥感图像中都能够有效地提取出图像中的各种端元.     

基于时频子空间分解的宽带线性调频信号DOA估计

针对具有时变方向向量的宽带线性调频信号,该文建立了基于短时Wigner-Ville分布(WVD)的空间时频分布矩阵,通过对各个空间时频矩阵的特征分解获得对应的信号子空间和噪声子空间,给出了基于时频子空间投影实现多个时频点综合估计信号DOA的算法。利用空间时频分布的前后向平滑解决了具有相同时频特性信号的均匀线阵DOA估计问题。算法不需要聚汇和插值等复杂的矩阵变换,精度较高,计算简便.仿真实验显示该算法性能显著优越于基于矩阵插值的宽带调频信号DOA估计算法.     

基于贝叶斯分类器的图像隐写分析

集成分类器是目前用于图像隐写分析的主流分类器。为提高集成分类器的检测精度,针对集成分类器基分类器组合方法过于简单,无法体现基分类器之间的内在联系,不能从整体上对结果进行判定的缺点,依据图像特征在集成分类器分类超平面上的投影值服从多维正态分布这一特性,提出了一种基于贝叶斯分类器的图像隐写分析算法。首先基于随机森林算法生成若干基分类器,然后计算类条件概率密度函数与先验概率并训练贝叶斯分类器,最后使用经过训练的贝叶斯分类器代替简单投票方法进行分类判决。算法的检测错误率比以往算法平均降低了1.6%,ROC曲线比简单投票方法更接近于左上角,即具有更高的检测率,AUC值平均增长约2.12%,并且训练时间仅有少量提高,最大提高约2.610s。可以有效提高集成分类器的检测精度。     

改进的OPAST算法及其在盲多用户检测中的应用

文中详细地介绍了正交投影子空间跟踪算法(OPAST),它是一种基于最优化问题的方法,保证了每次迭代时权向量的正交性,并具有和PAST算法一样的线性复杂度,以及与自然幂法(NP)一样的全局收敛性。然而将其应用于盲多用户检测时,在迭代一定次数后,会出现误码率突然增大现象,这就导致了算法性能的下降,为了解决这一问题,文中提出一种方法,并通过仿真结果,证明它是行之有效的。     

A novel steganalysis framework of heterogeneous images based on GMM clustering

The current steganalysis frameworks involve a large number of techniques for feature extraction and classification. However, one of their common defects is treating all images as equal, thus ignoring the variability of statistical properties of different images, which motivates us to propose a novel steganalysis framework based on Gaussian mixture model (GMM) clustering in the study, targeting at heterogeneous images with different texture complexity. There are two main improvements compared to the current steganalysis frameworks. First, in the training stage, the GMM clustering algorithm is exploited to classify the training samples into limited categories automatically, and then design corresponding steganalyzers for each category; second, in the testing stage, the posterior probability of testing samples belonging to each category is calculated, and the samples are submitted to the steganalyzers corresponding to the maximum posterior probability for test. Extensive experimental results aiming at least significant bit matching (LSBM) steganography and two adaptive steganography algorithms show that the proposed framework outperforms the steganalysis system that is directly trained on a mixed dataset, and also indicate that our framework exhibits better detection performance compared to the representative framework for using image contents in most circumstances and similar detection performance in few cases.     

A New Approach for Estimating the Number of Sources Under the Coexistence of Circular and Various Noncircular Sources

Estimation of source number is a fundamental problem of direction-of-arrival (DOA) estimation. In the problem of DOA estimation under the coexistence of circular and various noncircular signals, the source number should be estimated in order to distinguish the signal subspace from the noise subspace. Thus, a new method for source number estimation is proposed in this paper. Using the approach of k-means clustering, the projections of a one-dimensional reduced covariance matrix are divided into two categories. Then the signal subspace and the noise subspace are separated by the optimal classification boundary of those two categories so as to obtain the equivalent source number. Simulation results show that the proposed method has relatively better performance even in low SNR or in a colored noise environment.