Search

Text has become one of the most extensively used digital media in Internet, which provides steganography an effective carrier to realize confidential message hiding. Nowadays, generation-based linguistic steganography has made a significant breakthrough due to the progress of deep learning. However, previous methods based on recurrent neural network have two deviations including exposure bias and embedding deviation, which seriously destroys the security of steganography. In this paper, we propose a novel linguistic steganographic model based on adaptive probability distribution and generative adversarial network, which achieves the goal of hiding secret messages in the generated text while guaranteeing high security performance. First, the steganographic generator is trained by using generative adversarial network to effectively tackle the exposure bias, and then the candidate pool is obtained by a probability similarity function at each time step, which alleviates the embedding deviation through dynamically maintaining the diversity of probability distribution. Third, to further improve the security, a novel strategy that conducts information embedding during model training is put forward. We design various experiments from different aspects to verify the performance of the proposed model, including imperceptibility, statistical distribution, anti-steganalysis ability. The experimental results demonstrate that our proposed model outperforms the current state-of-the-art steganographic schemes.

With the extensive use of texts on social network, text steganography, which protects several sensitive messages by embedding secret data into normal texts, has attracted widespread attention. As an adversary, text steganalysis which reveals the existence of hidden messages is also important. Recently, Deep Neural Networks (DNNs) have led to significant improvements in text steganalysis. However, the deeper and wider DNNs cause the increase of inference time, which restricts the practicality of text steganalysis. In this paper, we propose an effective and real-time text steganalysis method based on multi-stage transfer learning to enhance inference efficiency and detection performance simultaneously. The experimental results show that the proposed text steganalysis method can outperform previously reported methods in terms of detection accuracy and inference efficiency.

Speech is one of the essential ways of communication. The study of speech steganography provides great value in information security. To improve imperceptibility and robustness of speech steganography, the characteristics of speech signals should be fully taken into account. In this paper, a robust speech steganographic scheme based on Singular Value Decomposition (SVD) and Modified Discrete Cosine Transform (MDCT) is proposed. Firstly, Voice Activity Detector (VAD) is used to detect voiced frames from speech signals, along with MDCT with Kaiser Bessel Derived (KBD) window being performed on each frame. Then the MDCT coefficients are selected from a certain frequency range and divided into a pair of segments. The two largest singular values of the paired segments are modified respectively according to their value difference to embed secret message. The thresholds are adaptively adjusted according to the largest singular values. Extensive experiments are carried out to compare the proposed method with three other methods from imperceptibility, robustness, capacity, and security. The experimental results show that under the simulation parameters β = 320, Nk = 58, fl = 100 Hz, fh = 3 kHz, and α = 0.61, the proposed method has striking advantages to resist common robust attacks and the state-of-the-art steganalysis attacks while maintaining good imperceptibility.

The conventional classification models implicitly assume that the distributions of data employed for training and test are identical. However, the assumption is rarely valid in many practical applications. In order to alleviate the difference between the distributions of the training and test sets, in this paper, we propose a regularized subspace learning framework based on the low-rank representation technique for unsupervised domain adaptation. Specifically, we introduce a regularization term of the subspace projection matrix to deal with the ill-conditioned problem and obtain a unique numerical solution. Meanwhile, we impose a structured sparsity-inducing regularizer on the error term so that the proposed method can filter out the outlier information, and therefore improve the performance. The extensive comparison experiments on benchmark data sets demonstrate the effectiveness of the proposed method.

With the emergence of the generation-based steganography, the traditional text steganalysis methods show the unsatisfactory detection performance as the manually extracted features are simple and non-universal. The recently proposed deep learning-based text steganalysis methods can obtain the great detection accuracy by extracting the high-level features. In this letter, a hybrid text steganalysis method (R-BILSTM-C) is proposed through combining the advantages of Bidirectional Long Short Term Memory Recurrent Neural Network (Bi-LSTM) and Convolutional Neural Network (CNN). The proposed method can efficiently capture both local features and long-term semantic information from text to improve the detection accuracy. In the proposed method, the Bi-LSTM architecture is used to capture the long-term semantic information of texts. And the asymmetric convolution kernels with different sizes are applied to extract the local relationship between words. In addition, the high dimensional semantic feature space is visualized. Experimental results show that the proposed method adapts to the different steganographic algorithms efficiently, and achieves the comparable or superior detection performance for the various sentence lengths compared with other state-of-the-art text steganalysis methods.

The prevailing text steganalysis methods detect steganographic communication by extracting hand-crafted features and classifying them using SVM. However, these features are designed based on the statistical changes caused by steganography, thus they are difficult to adapt to different kinds of embedding algorithms and the detection performance is heavily dependent on the text size. In this letter, we propose a novel text steganalysis model based on convolutional neural network, which is able to capture complex dependencies and learn feature representations automatically from the texts. First, we use a word embedding layer to extract the semantic and syntax feature of words. Second, the rectangular convolution kernels with different sizes are used to learn the sentence features. To further improve the performance, we present a decision strategy for detecting the long texts. Experimental results show that the proposed method can effectively detect different kinds of text steganographic algorithms and achieve comparable or superior performance for a wide variety of text sizes compared with the previous methods.

The matrix-based multi-view feature selection, which can integrate the information of multiple views for selecting representative features, has attracted wide attention in recent years. In this paper, we propose a novel supervised sharing multi-view feature selection method. The proposed method makes all views share a common penalty that regresses samples to their labels. Meanwhile, it adopts the structured sparsity-inducing norm to implement sparsity for each view. The proposed method considers not only the complementary of different views, but also the specificity of each view instead of concatenating all views into high-dimensional vectors. In addition, the proposed model can be decomposed into N small scale subproblems (where N is the number of views) and solved efficiently via Alternating Direction Method of Multipliers (ADMM), especially for high-dimensional large scale data sets. The comparison experiments with several state-of-the-art feature selection methods show the effectiveness of the proposed method.

Currently, various agricultural image classification tasks are carried out on high-resolution images. However, in some cases, we cannot get enough high-resolution images for classification, which significantly affects classification performance. In this paper, we design a crop disease classification network based on Enhanced Super-Resolution Generative adversarial networks (ESRGAN) when only an insufficient number of low-resolution target images are available. First, ESRGAN is used to recover super-resolution crop images from low-resolution images. Transfer learning is applied in model training to compensate for the lack of training samples. Then, we test the performance of the generated super-resolution images in crop disease classification task. Extensive experiments show that using the fine-tuned ESRGAN model can recover realistic crop information and improve the accuracy of crop disease classification, compared with the other four image super-resolution methods.

This article presents the linear Proximal Support Vector Machine (PSVM) to the image steganalysis, and further generates a very efficient method called PSVM-LSMR through implementing PSVM by the state-of-the-art optimization method Least Square Minimum-Residual (LSMR). Also, motivated by extreme learning machine (ELM), a nonlinear algorithm PSVM-ELM is proposed for the image steganalysis. It is shown by the experiments with the wide stego schemes and rich steganalysis feature sets in both the spatial and JPEG domains that the PSVM can achieve comparable performance with Fisher Linear Discriminant (FLD) and ridge regression, and its computational time is far more less than that of them on large feature sets. The PSVM-LSMR is comparable to Ridge Regression implemented by LSMR (RR-LSMR), and both of them require the least computational time among all the competitions when dealing with medium or large feature sets. The nonlinear PSVM-ELM performs comparably or even better than FLD and ridge regression for the spatial domain steganographic schemes, and its computational time is apparently less than that of them on large feature sets.

In applications, using features obtained from multiple views to describe objects has become popular because multiple views contain much more information than the single view. As the dimensions of the data sets are high, which may cause expensive time consumption and memory space, how to identify the representative views and features becomes a crucial problem. Multi-view feature selection that can integrate multiple views to select important and relevant features to improve performance has attracted more and more attentions in recent years. Previous supervised multi-view feature selection methods usually establish the models by concatenating multiple views into long vectors. However, this concatenation is not physically meaningful and implies that different views play the similar roles for specific tasks. In this paper, we propose a novel supervised multi-view feature selection method based on the weighted hinge loss (WHMVFS) that can learn the corresponding weight for each view and implement sparsity from the group and individual point of views under the structured sparsity framework. The newly proposed multi-view weighted hinge loss penalty not only has the ability to select more discriminative features for classification, but also can make the involved optimization problem be decomposed into several small scale subproblems, which can be easily solved by an iterative algorithm, and the convergence of the iterative algorithm is also proved. Experimental results conducted on real-world data sets show the effectiveness of the proposed method.

The Android permission mechanism prevents malicious application from accessing the mobile multimedia data and invoking the sensitive API. However, there are still lots of deficiencies in the current permission management, which results in the permission mechanism being unable to protect users’ private data properly. In this paper, a dynamic management scheme of Android permission based on machine learning is proposed to solve the problem of the existing permission mechanism. In order to accomplish the dynamic management, the proposed scheme maintains a dynamic permission management database which records the state of permissions for each application. Only the permission which is granted state in the database can be used in this application. In the whole process, the scheme first classifies the application by means of machine learning, then retrieves the corresponding permission information from databases, and issues the dangerous permission warning to users. Finally, the scheme updates the dynamic management database according to the users’ decisions. Through this scheme, users can prevent malicious behaviour of accessing private data and invoking sensitive API in time. The solution increases the flexibility of permission management and improves the security and reliability of multimedia data in Android devices.

The prevailing steganalysis detector trained by a source is used to recognize images from another different source, the detection accuracy typically drops owing to the mismatch between the two sources. In contrast to previous mismatched steganalysis methods, in this paper, we develop an unsupervised subspace learning-based method which has some differences from the ones common used in mismatched steganalysis. The proposed method employs low-rank and sparse constraints on the reconstruction coefficient matrix to maintain the global and local structures of the data. In this way, we can obtain new feature representations so that the feature distributions of the training and test data are close. We further promote the performance of the proposed method by employing the l2,1-norm on the error matrix. Comprehensive experiments on the JPEG mismatched steganalysis are conducted, and the experimental results show that the proposed method can improve the detection accuracy.

The convolutional neural network (CNN) based image steganalyzers have evolved remarkably over the past few years, and designing suitable CNN structures has been currently the fundamental method to improve the detection accuracy. However, CNN-based steganalyzer with the universal activation functions of the computer vision (CV) field barely achieves significant performance improvement. Therefore, a dedicated activation function is required to improve the detection performance for image steganalysis. In this paper, we propose a point-wise parametric rectified unit (PW-PReLU) which has different adaptively learnable parameters for each pixel of the negative inputs to facilitate the representation capacity of the activated feature maps. Then, in order to further boost the detection accuracy, the feature fusion is realized by the concatenation operation in the first layer. Based on the above components, an optimized CNN-based steganalyzer is proposed for spatial image steganalysis. The results of comparative experiments demonstrate that the proposed network can detect the state-of-the-art spatial steganographic algorithms with better performance than the previous steganalyzers on the 512 \(\times \) 512 BOSSbase_1.01 dataset and the resized 256 \(\times \) 256 union dataset of BOSSbase_1.01 and BOWS2.

Image steganalysis is a very important research topic in the field of information security. The existing feature based image steganalysis methods have achieved the appealing performance. The performance of them greatly depends on the quality of the hand-crafted steganalysis feature vectors, such as Cartesian Calibration PEV (CC-PEV), DCT Residuals (DCTR), and so on. However, these feature vectors may contain some redundant elements that will reduce the discrimination power and increase the computation cost. In this paper, a novel feature selection model is proposed for JPEG image steganalysis. Specifically, the proposed model imposes an \(l_{2,1}\)-structural constraint on the projection matrix for feature selection. Further, to make the model insensitive to noises and outliers, a capped \(l_2\)-norm based loss function is adopted. Moreover, a graph-based manifold regularization term which exploits the intrinsic local geometric structure of the data is added into the objective function to select the effective feature elements. Finally, an alternately iterative optimization algorithm with proven convergence is given to solve the proposed model. The extensive experiments on three state-of-the-art JPEG steganographic algorithms with 0.1 and 0.2 embedding rates and two JPEG quality factors show that the proposed model can effectively remove some irrelevant and redundant elements meanwhile retaining high detection accuracy.

In recent years, manufacturing execution system (MES) has been widely used in the manufacturing industry. MES plays an important role in supporting enterprise workshop management. However, generalized MES cannot be applied well in steel structure enterprises due to the particularity and complexity of steel structure enterprises. This paper takes into account the production characteristics and needs of steel structure enterprises and an MES structure based on cyber-physical system (CPS) is proposed. This architecture includes the physical layer, the network layer and the application layer. The cloud platform-based MES system was applied to a steel structure enterprise in Changchun, China, and the functions of each module were introduced in detail. The MES can effectively improve the informatization level of the enterprise, creating a transparent, informative and intelligent production environment for steel structure enterprises. The developed MES can improve resource utilization, reduce enterprise costs, improve product quality, and lay the foundation for steel structure enterprises to found intelligent factories.

Due to the limited varieties and sizes of existing public hyperspectral image (HSI) datasets, the classification accuracies are higher than 99% with convolutional neural networks (CNNs). In this paper, we presented a new HSI dataset named Shandong Feicheng, whose size and pixel quantity are much larger. It also has a larger intra-class variance and a smaller inter-class variance. State-of-the-art methods were compared on it to verify its diversity. Otherwise, to reduce overfitting caused by the imbalance between high dimension and small quantity of labeled HSI data, existing CNNs for HSI classification are relatively shallow and suffer from low capacity of feature learning. To solve this problem, we proposed an HSI classification framework named deep residual spectral spatial setwork (DRSSN). By using shortcut connection structure, which is an asymmetry structure, DRSSN can be deeper to extract features with better discrimination. In addition, to alleviate insufficient training caused by unbalanced sample sizes between easily and hard classified samples, we proposed a novel training loss function named sample balanced loss, which allocated weights to the losses of samples according to their prediction confidence. Experimental results on two popular datasets and our proposed dataset showed that our proposed network could provide competitive results compared with state-of-the-art methods.

With industrial internet increasingly used in manufacturing industry, the traditional manufacturing industry can no longer meet the needs of the development of manufacturing and producing, which brought both an opportunity and a huge challenge for small and medium-sized manufacturing enterprises. After investigations and consultations with many small and medium-sized enterprise, developed a set of production management systems for small and medium-sized enterprises. The system includes three layers: physical layer, network layer and system layer. At the physical layer, field data is collected through photoelectric sensors and a model is established to found high-precision sensor node method. At the network layer, ZigBee wireless transmission network protocol is used to develop a wireless transmission module with strong anti-interference ability and data confidentiality. At the system layer, a production management system is developed according to characteristics of mobile terminals and PC in order to achieve equipment failure repair, alarm message push and other functions in the mobile terminal and to realize visual board and other functions. The system and technologies proposed is important for information-based manufacturing industries.

Along with the rapid development of science and technology, power electronic technology has developed rapidly. Especially in recent years, the application of power electronic technology can be seen in many fields. In the smart grid with information, digitization, automation and interaction as the development goal, power electronics technology has become one of the key technologies, is an important technical means to build smart grid. Reasonable and efficient use of power electronic technology in smart grid can effectively improve the security, stability and economic operation level of power grid, and can greatly improve the power grid transportation power technology support. This article through the power electronic technology used in the smart grid to carry on the analysis, a comprehensive analysis of power electronics in smart grid applications of facts, HVDC, intelligent switch technology, high frequency technology in the smart grid, and points out its future development priorities, so as to guarantee the safety and stability of more scientific and the economic operation of the power grid.

Employee Assistance Program (EAP), originated in the United States, is an important method of human resource management, which is widely applied in the enterprises of developed countries. With the rapid development of Chinese economy, more and more hotel managers recognize that there much be relations between employees' body and mental health and hotel performance. At present, problems such as how to improve employees' health condition, increase their satisfaction and reduce employee turnover rate have been ahead of hotel managers. This paper puts forward six effective ways of EAP application among hotel staff and five key issues that have to be noted.

Reactive power cannot be transmitted over long distances and voltage stability is considered as more of local problem rather than a global problem. So this paper presents a new method of combining network partitioning and voltage stability sensitivity index for optimal allocation of reactive sources to enhance voltage stability. The concept of electrical distance is introduced for network partitioning, partitioning a power network into strongly connected sub-networks. By applying this approach, the reactive sources can be allocated most reasonable, reactive compensation balance in every zone. The simulation results indicated that the method is practicable and effective to improve the voltage stability of power systems.