Search

Fuzzy target detection as an important task to reflect the detection ability of underwater robot, the artificial target recognition based on the image taken by underwater robot has been widely concerned. However, there is no open standard fuzzy underwater image data set, and in the harsh deep-water fuzzy environment, it is difficult to collect large-scale marked underwater fuzzy optical images. At the same time, it is also hoped that the detection model has the ability to learn quickly from small samples in the case of as few samples as possible. Therefore, combining depth learning and transfer learning, a new method based on improved SSD and transfer learning is proposed. Firstly, we design a more accurate SSD network (underwater SSD) which is suitable for fuzzy underwater target detection. The features extracted from the detection network are highly representative. Secondly, we use the Transfer learning method to train the underwater SSD network, which can only use the tags in the air to identify fuzzy underwater objects, and have strong robustness in both the air and fuzzy underwater imaging modes. Finally, soft NMS is used to detect the target. The experimental results of the simulation data show that the algorithm not only overcomes the difficulties of the known data set of underwater target, but also effectively improves the accuracy of underwater target detection compared with the traditional deep learning method, reaching 82.31%, showing better detection performance.

Facial expression recognition plays an increasingly important role in monitoring the mental state of individual soldiers. Aiming at the problem of low accuracy rate by traditional facial expression recognition method, this paper proposes an improved shallow residual network facial expression recognition algorithm. Based on ResNet which is the main target detection framework of existing deep learning, a Res-HyperNet structure is proposed to learn more shallow features. First, the shallow features are aggregated and compressed into one space. To improve the recognition effect, this paper sent the high-level features which have more useful semantic information to a deeper convolution layer through a shortcut to perform fusion calculation. Three public datasets (CK+, JAFFE, Oulu-CASIA) and self-made datasets are used for accuracy rate experiments. Compared with the existing mainstream facial expression recognition method, the improved shallow residual network algorithm can raise the accuracy rate up to 3%.

Inpainting, originally designed in computer vision to reconstruct lost or deteriorated parts of images and videos, has been used for image tampering, including region filling and object removal to alter the truth. While several types of tampering including copy-move and seam carving forgery can now be successfully exposed in image forensics, there has been very little study to tackle inpainting forgery in JPEG images, the detection of which is extremely challenging due to the post-recompression attacks performed to cover or compromise original inpainting traces. To date, there is no effective way to detect inpainting image forgery under combined recompression attacks. To fill such a gap in image forensics and reveal inpainting forgery from the post-recompression attacks in JPEG images, we propose in this paper an approach that begins with large feature mining in discrete transform domain, ensemble learning is then applied to deal with the high feature dimensionality and to prevent the overfitting that generally happens to some regular classifiers under high feature dimensions. Our study shows the proposed approach effectively exposes inpainting forgery under post recompression attacks, especially, it noticeably improves the detection accuracy while the recompression quality is lower than the original JPEG image quality, and thus bridges a gap in image forgery detection.

Digital multimedia forensics is an emerging field that has important applications in law enforcement and protection of public safety and national security. In digital imaging, JPEG is the most popular lossy compression standard and JPEG images are ubiquitous. Today's digital techniques make it easy to tamper JPEG images without leaving any visible clues. Furthermore, most image tampering involves JPEG double compression, it heightens the need for accurate analysis of JPEG double compression in image forensics. In this paper, to improve the detection of JPEG double compression, we transplant the neighboring joint density features, which were designed for JPEG steganalysis, and merge the joint density features with marginal density features in DCT domain as the detector for learning classifiers. Experimental results indicate that the proposed method improves the detection performance. We also study the relationship among compression factor, image complexity, and detection accuracy, which has not been comprehensively analyzed before. The results show that a complete evaluation of the detection performance of different algorithms should necessarily include image complexity as well as the double compression quality factor. In addition to JPEG double compression, the identification of image capture source is an interesting topic in image forensics. Mobile handsets are widely used for spontaneous photo capture because they are typically carried by their users at all times. In the imaging device market, smartphone adoption is currently exploding and megapixel smartphones pose a threat to the traditional digital cameras. While smartphone images are widely disseminated, the manipulation of images is also easily performed with various photo editing tools. Accordingly, the authentication of smartphone images and the identification of post-capture manipulation are of significant interest in digital forensics. Following the success of our previous work in JPEG double compression detection, we conducted a study to identify smartphone source and post-capture manipulation by utilizing marginal density and neighboring joint density features together. Experimental results show that our method is highly promising for identifying both smartphone source and manipulations. Finally, our study also indicates that applying unsupervised clustering and supervised classification together leads to improvement in identifying smartphone sources and manipulations and thus provides a means to address the complexity issue of the intentional post-capture manipulation on smartphone images.

The prosperity of mobile devices have been rapidly and drastically reforming the use pattern and of user habits with computing devices. Android, the most popular mobile operating system, has a privilege-separated security system through a sophisticated permission control mechanism. Android Apps need to request permissions to access sensitive personal data and system resources, but empirical studies have found that various types of malicious software could obtain permissions and attack systems and applications by deceiving users and the security mechanism. In this paper, we propose a novel machine learning approach to detect malware by mining the patterns of Permissions and API Function Calls acquired and used by Android Apps. Based on static analysis of source code and resource files of Android Apps, binary and numerical features are extracted for qualitative and quantitative evaluation. Feature selection methods are applied to reduce the feature dimension and enhance the efficiency. Different machine learning methods, including Support Vector Machines, Random Forest and Neural Networks, are applied and compared in classification. The experimental results show that the proposed approach delivers accurate detection of Android malware. We deem that the proposed approach could help raise users' awareness of potential risks and mitigate malware threats for Android devices.

This article presents a second-order derivative-based audio steganalysis. First, Mel-cepstrum coefficients and Markov transition features from the second-order derivative of the audio signal are extracted; a support vector machine is then applied to the features for discovering the existence of hidden data in digital audio streams. Also, the relation between audio signal complexity and steganography detection accuracy, which is an issue relevant to audio steganalysis performance evaluation but so far has not been explored, is analyzed experimentally. Results demonstrate that, in comparison with a recently proposed signal stream-based Mel-cepstrum method, the second-order derivative-based audio steganalysis method gains a considerable advantage under all categories of signal complexity--especially for audio streams with high signal complexity, which are generally the most challenging for steganalysis-and thereby significantly improves the state of the art in audio steganalysis.

The threat posed by hackers, spies, terrorists, and criminals, etc. using steganography for stealthy communications and other illegal purposes is a serious concern of cyber security. Several steganographic systems that have been developed and made readily available utilize JPEG images as carriers. Due to the popularity of JPEG images on the Internet, effective steganalysis techniques are called for to counter the threat of JPEG steganography. In this article, we propose a new approach based on feature mining on the discrete cosine transform (DCT) domain and machine learning for steganalysis of JPEG images. First, neighboring joint density features on both intra-block and inter-block are extracted from the DCT coefficient array and the absolute array, respectively; then a support vector machine (SVM) is applied to the features for detection. An evolving neural-fuzzy inference system is employed to predict the hiding amount in JPEG steganograms. We also adopt a feature selection method of support vector machine recursive feature elimination to reduce the number of features. Experimental results show that, in detecting several JPEG-based steganographic systems, our method prominently outperforms the well-known Markov-process based approach.

MPEG-1 Audio Layer 3, more commonly referred to as MP3, is a popular audio format for consumer audio storage and a de facto standard of digital audio compression for the transfer and playback of music on digital audio players. MP3 audio forgery manipulations generally uncompress a MP3 file, tamper with the file in the temporal domain, and then compress the doctored audio file back into MP3 format. If the compression quality of doctored MP3 file is different from the quality of original MP3 file, the doctored MP3 file is said to have undergone double MP3 compression. Although double MP3 compression does not prove a malicious tampering, it is evidence of manipulation and thus may warrant further forensic analysis since, e.g., faked MP3 files can be generated by using double MP3 compression at a higher bit-rate for the second compression to claim a higher quality of the audio files. To detect double MP3 compression, in this paper, we extract the statistical features on the modified discrete cosine transform and apply a support vector machine to the extracted features for classification. Experimental results show that our designed method is highly effective for detecting faked MP3 files. Our study also indicates that the detection performance is closely related to the bit-rate of the first-time MP3 encoding and the bit-rate of the second-time MP3 encoding.

Steganography secretly embeds additional information in digital products, the potential for covert dissemination of malicious software, mobile code, or information is great. To combat the threat posed by steganography, steganalysis aims at the exposure of the stealthy communication. In this paper, a new scheme is proposed for steganalysis of JPEG images, which, being the most common image format, is believed to be widely used for steganography purposes as there are many free or commercial tools for producing steganography using JPEG covers. First, a recently proposed Markov approach [27] is expanded to the inter-block of the discrete cosine transform (DCT) and to the discrete wavelet transform (DWT). The features on the joint distributions of the transform coefficients and the features on the polynomial fitting errors of the histogram of the DCT coefficients are also extracted. All features are called original ExPanded Features (EPF). Next, the EPF features are extracted from the calibrated version; these are called reference EPF features. The difference between the original and the reference EPF features is calculated, and then the original EPF features and the difference are merged to form the feature vector for classification. To handle the large number of developed features, the feature selection method of support vector machine recursive feature elimination (SVM-RFE) and a method of multi-class support vector machine recursive feature elimination (MSVM-RFE) are used to select features for binary classification and multi-class classification, respectively. Finally, support vector machines are applied to the selected features for detecting stego-images. Experimental results show that, in comparison to the Markov approach [27], this new scheme remarkably improves the detection performance on several JPEG-based steganographic systems, including JPHS, CryptoBola, F5, Steghide, and Model based steganography.

To improve a recently developed mel-cepstrum audio steganalysis method, we present in this paper a method based on Fourier spectrum statistics and mel-cepstrum coefficients, derived from the second-order derivative of the audio signal. Specifically, the statistics of the high-frequency spectrum and the mel-cepstrum coefficients of the second-order derivative are extracted for use in detecting audio steganography. We also design a wavelet-based spectrum and mel-cepstrum audio steganalysis. By applying support vector machines to these features, unadulterated carrier signals (without hidden data) and the steganograms (carrying covert data) are successfully discriminated. Experimental results show that proposed derivative-based and wavelet-based approaches remarkably improve the detection accuracy. Between the two new methods, the derivative-based approach generally delivers a better performance.

With the booming of smart phone and high-speed wireless networks in recent years, applications and data have been shifting from desktop to mobile devices at a vigorous pace. Although mobile computing provides great convenience in daily life, it becomes vulnerable to various types of emerging attacks. User authentication plays an indispensable role in protecting computer systems and applications, but the development of touch screen hardware and user habit change post requirements for new authentication methods for mobile and tablets devices. In this paper, we present a robust user authentication scheme using both static and dynamic features of touch gestures. We take advantage of the pressure sensitivity of multi-touch screens to obtain irreproducible biometric patterns. Discriminative features such as distance, angle, and pressure are extracted from the touch-point data, and used in statistical analysis for verification. We tested our scheme in a variety of experiments that involved multiple volunteers to perform various gestures. The analysis of experimental results and user feedback indicate the proposed scheme delivers comprehensive measurements and accurate pattern classification for touch gestures. Based on these results, we conclude that the proposed scheme overcomes the limitations of the existing user authentication methods, and shows great potential to provide robust protection against unauthorized access.

Keywords: RCA (revealed comparative advantage), maxproximity, product space Abstract. This paper emphasize "maxproximity" to define the maximum proximity between a product with RCA 1, then calculate the relationship between them. We take China as an example to verify the feasibility on maxproximity, simulate the product structure evolution with the rules we summarize, and construct a product space network to demonstrate the process. We draw a conclusion that country specialize the products with high complexity can produce other kinds of goods easily, because the country has the product knowledge more diversified.

With the booming of smartphone and high-speed wireless networks in recent years, audio streaming and sharing become convenient and inexpensive, so that digital media is gradually replacing physical media. This trend has also led to more attacks to digital audio and its application. MP3 double compression, achieved by decompressing and recompressing audio to a different compression ratio, is a typical manipulation of audio for malicious purposes. In this paper, we propose an approach for detecting both up-transcoded and down-transcoded MP3 audio files and revealing the real compression quality based on statistical patterns extracted from quantized MDCT coefficients and their derivatives. To minimize the false prediction caused by individual characteristics of diversified audio clips, we generated reference audio signals by recompressing and calibrating the audio, and measured the differences between signal-based and reference-based features. Support vector machines and dynamic evolving neural-fuzzy inference systems were applied for binary and multi-class classifications. The experimental results show that our approach effectively detects MP3 double compression and exposes the audio processing history for digital forensics.

As smartphones are being widely used in daily lives, the images captured by smartphones become ubiquitous and may be used for legal purposes. Accordingly, the authentication of smartphone images and the identification of post-capture manipulation are of significant interest in digital forensics. In this paper, we propose a method to determine the smartphone camera source of a particular image and operations that may have been performed on that image. We first take images using different smartphones and purposely manipulate the images, including different combinations of double JPEG compression, cropping, and rescaling. Then, we extract the marginal density in low frequency coordinates and neighboring joint density features on intra-block and inter-block as features. Finally, we employ a support vector machine to identify the smartphone source as well as to reveal the operations. Experimental results show that our method is very promising for identifying both smartphone source and manipulations. Our study also indicates that applying unsupervised clustering and supervised classification together (clustering first, followed by classification) leads to improvement in identifying smartphone sources and manipulations and thus provides a means to address the complexity issue of intentional manipulation.

Steganography, the ancient art for secretive communications, has revived on the Internet by way of hiding secret data, in completely imperceptible manners, into a digital file. Thus, the steganography has created a serious threat to cyber security due to the covert channel it provides that can be readily exploited for various illegal purposes. Likewise, image tampering or forgery, which has been greatly facilitated and proliferated by photo processing tools, is increasingly causing problems concerning the authenticity of digital images. JPEG images constitute one of the most popular media on the Internet; yet they can be easily used for the steganography as well as easily tampered by, e.g., removing, adding, or splicing objects without leaving any clues. Therefore, there is a critical need to develop reliable methods for steganalysis (analysis of multimedia for the steganography) and for forgery detection in JPEG images to serve applications in national security, law enforcement, cybercrime fighting, digital forensics, and network security, etc. This article presents some recent results on detecting JPEG steganograms, doubly compressed JPEG images, and resized JPEG images based on a unified framework of feature mining and pattern recognition approaches. At first, the neighboring joint density features and marginal density features of the DCT coefficients of the JPEG image are extracted; then learning classifiers are applied to the features for the detection. Experimental results indicate that the method prominently improves the detection performances in JPEG images when compared to a previously well-studied method. Also, it is demonstrated that detection performance deteriorates with increasing image complexity; hence, a complete evaluation of the detection performance of different algorithms should include image complexity—in addition to other relevant factors such as hiding ratio or compression ratio—as a significant and independent parameter.

MP3 is the most popular format for audio storage and a de facto standard of digital audio compression for the transfer and playback. The flexibility of compression ratio of MP3 coding enables users to choose their customized configuration in the trade-off between file size and quality. Double MP3 compression often occurs in audio forgery, steganography and quality faking by transcoding an MP3 audio to a different compression ratio. To detect double MP3 compression, in this paper, we extract the statistical features on the modified discrete cosine transform, and apply support vector machines and a dynamic evolving neuron-fuzzy inference system to the extracted features for classification. Experimental results show that our method effectively and accurately detects double MP3 compression for both up-transcoded and down-transcoded MP3 files. Our study also indicates the potential for mining the audio processing history for forensic purposes.

As a serious concern of information security, steganography provides a covert communication channel for cyber-terrorism and cyber-crime. The widespread use enables MP3 compressed audio to become an excellent carrier for audio steganography on the Internet. Since embedding capacity is an important measure to evaluate the performance of steganographic systems, and embedding ratio is commonly used when comparing the accuracy of different steganalysis algorithms. In this paper, we present a scheme to predict embedding strength based on feature mining in MDCT transform domain. We generate reference signals by reversing and repeating quantification process, and derive the reference based accumulative features from the difference between source signal and reference signal. Finally, a dynamic evolving neuron-fuzzy inference system is applied to predict embedding strength of MP3 compressed audio. Experimental results show that our approach obtains good prediction of the embedding strength in the steganograms.

In this paper, we present a novel stream data mining for audio steganalysis, based on second order derivative of audio streams. We extract Mel-cepstrum coefficients and Markov transition features on the second order derivative, a support vector machine is applied to the features for discovery of the existence of covert message in digital audios. We also explore the relation between signal complexity and detection performance on digital audios, which has not been studied previously. Our study shows that, in comparison with a recently proposed signal stream based Mel-cepstrum steganalysis, our method prominently improves the detection performance, which is not only related to information-hiding ratio but also signal complexity. Generally speaking, signal stream based Mel-cepstrum audio steganalysis performs well in steganalysis of audios with low signal complexity; it does not work so well on audios with high signal complexity. Our stream mining approach for audio steganalysis gains significant advantage in each category of signal complexity - especially in audios with high signal complexity, and thus improves the state of the art in audio steganalysis.

Detection of information-hiding in JPEG images is actively delivered in steganalysis community due to the fact that JPEG is a widely used compression standard and several steganographic systems have been designed for covert communication in JPEG images. In this paper, we propose a novel method of JPEG steganalysis. Based on an observation of bi-variate generalized Gaussian distribution in Discrete Cosine Transform (DCT) domain, neighboring joint density features on both intra-block and inter-block are extracted. Support Vector Machines (SVMs) are applied for detection. Experimental results indicate that this new method prominently improves a current art of steganalysis in detecting several steganographic systems in JPEG images. Our study also shows that it is more accurate to evaluate the detection performance in terms of both image complexity and information hiding ratio.

Introduction In the classification of Mass Spectrometry (MS) proteomics data, peak detection, feature selection, and learning classifiers are critical to classification accuracy. To better understand which methods are more accurate when classifying data, some publicly available peak detection algorithms for Matrix assisted Laser Desorption Ionization Mass Spectrometry (MALDI-MS) data were recently compared; however, the issue of different feature selection methods and different classification models as they relate to classification performance has not been addressed. With the application of intelligent computing, much progress has been made in the development of feature selection methods and learning classifiers for the analysis of high-throughput biological data. The main objective of this paper is to compare the methods of feature selection and different learning classifiers when applied to MALDI-MS data and to provide a subsequent reference for the analysis of MS proteomics data. Results We compared a well-known method of feature selection, Support Vector Machine Recursive Feature Elimination (SVMRFE), and a recently developed method, Gradient based Leave-one-out Gene Selection (GLGS) that effectively performs microarray data analysis. We also compared several learning classifiers including K-Nearest Neighbor Classifier (KNNC), Naïve Bayes Classifier (NBC), Nearest Mean Scaled Classifier (NMSC), uncorrelated normal based quadratic Bayes Classifier recorded as UDC, Support Vector Machines, and a distance metric learning for Large Margin Nearest Neighbor classifier (LMNN) based on Mahanalobis distance. To compare, we conducted a comprehensive experimental study using three types of MALDI-MS data. Conclusion Regarding feature selection, SVMRFE outperformed GLGS in classification. As for the learning classifiers, when classification models derived from the best training were compared, SVMs performed the best with respect to the expected testing accuracy. However, the distance metric learning LMNN outperformed SVMs and other classifiers on evaluating the best testing. In such cases, the optimum classification model based on LMNN is worth investigating for future study.

In this article, we propose a scheme for detecting hidden messages in compressed audio files produced by MP3Stego, as our literature search has found no previous work on successful steganalysis of MP3Stego. We extract moment statistical features on the second derivatives, as well as Markov transition features and neighboring joint density of the MDCT coefficients based on each specific frequency band on MPEG-1 Audio Layer 3. A support vector machine is applied to different feature sets for classification. Experimental results show that our approach is successful to discriminate MP3 covers and the steganograms generated by using MP3Stego.

MP3 allows a high compression ratio while providing high fidelity. As it has become one of the most popular digital audio formats, MP3 is also conceivably a most utilized carrier for audio steganography, therefore, MP3 steganalysis is a topic deserving attention. In this paper, we propose a scheme for steganalysis of MP3Stego based on feature mining and pattern recognition techniques. We first extract the moment statistical features of GGD shape parameters of the MDCT sub-band coefficients, as well as the moment statistical features, neighboring joint densities, and Markov transition features of the second order derivatives of the MDCT coefficients on MPEG-1 Audio Layer 3. Support Vector Machines (SVM) are applied to these features for detection. Experimental results show that our method can successfully discriminate the steganograms created by using MP3stego from their MP3 covers, even with fairly low embedding ratio.

In the General Packet Radio Service (GPRS) network, data packets from services like mobility management and “push-to-talk” (PTT) are sensitive to delay. During the channel allocation process, the delay-sensitive packets should have higher priority than other packets. Previously priority queue and Uplink State Flag Channel Allocation (USFCA) have been used to reduce the transmission delay of priority packets. In this paper, we study the performance of channel allocation for priority packet in the GPRS network with multi-slot capability. The results show that the transmission delay of priority packet can be further reduced by assigning multiple channels to priority packet in addition to the use of priority queue and USFCA.

In this paper, we propose an audio steganalysis method called reference based Fourier Spectrum Steganalysis. The mean values and the standard deviations of the high frequency spectrum of the second and high order derivatives are extracted from the testing signals and the reference versions. A Support Vector Machine (SVM) is employed to discriminate the unadulterated carrier signals and the steganograms wherein covert messages were embedded. Experimental results show that our method delivers very good performance and holds great promise for effective detection of steganograms produced by Hide4PGP, Invisible Secrets, S-tools4 and Steghide.

In this article, we propose a steganalysis method for detecting the presence of information-hiding behavior in wav audios. We extract the neighboring joint distribution features and the Markov features of the second order derivative, and combine these features with the error response by randomly modifying the least significant bit, then apply learning machines to the features for distinguishing the stegoaudios from cover videos. Experimental results show that our method performs well in steganalysis of the audio stegograms that are produced by using Hide4PGP, Invisible Secrets and S-tools4.

Background Microarray data have a high dimension of variables and a small sample size. In microarray data analyses, two important issues are how to choose genes, which provide reliable and good prediction for disease status, and how to determine the final gene set that is best for classification. Associations among genetic markers mean one can exploit information redundancy to potentially reduce classification cost in terms of time and money. Results To deal with redundant information and improve classification, we propose a gene selection method, Recursive Feature Addition, which combines supervised learning and statistical similarity measures. To determine the final optimal gene set for prediction and classification, we propose an algorithm, Lagging Prediction Peephole Optimization. By using six benchmark microarray gene expression data sets, we compared Recursive Feature Addition with recently developed gene selection methods: Support Vector Machine Recursive Feature Elimination, Leave-One-Out Calculation Sequential Forward Selection and several others. Conclusions On average, with the use of popular learning machines including Nearest Mean Scaled Classifier, Support Vector Machine, Naive Bayes Classifier and Random Forest, Recursive Feature Addition outperformed other methods. Our studies also showed that Lagging Prediction Peephole Optimization is superior to random strategy; Recursive Feature Addition with Lagging Prediction Peephole Optimization obtained better testing accuracies than the gene selection method varSelRF.

Mass spectrometry has become a widely used measurement in proteomics research. High dimensionality of features and small dataset are two major limitations hindering the accuracy of classification in mass spectrum data analysis; consequently, to obtain good results, the issues of feature extraction and feature selection are especially important. The quality of the feature set determines the reliability of the prediction of disease status. A well-known approach is to detect peak values and then apply support vector machine recursive feature elimination (SVMRFE) to choose feature sets for classification. In this paper, we apply a distance metric learning to classify proteomics mass spectrometry data. Experimental results show that distance metric learning can successfully be applied to the classification of proteomics data and the results are comparable to or better than, the best results by applying SVM to the feature sets chosen with the use of SVMRFE. We also perform feature reduction using manifold learning and experimental results indicate its promising potential in this application.