Search

Image steganography is the art and science of secure communication by concealing information within digital images. In recent years, the techniques of steganographic cost learning have developed rapidly. Although the existing methods can learn satisfactory additive costs, the interplay of different pixels’ embedding impacts has not been considered, so the potential of learning may not be fully exploited. To overcome this limitation, in this paper, a reinforcement learning paradigm called JoPoL (joint policy learning) is proposed to extend the idea of additive cost learning to a non-additive situation. JoPoL aims to capture the interactions within pixel blocks by defining embedding policies and evaluating contributions of embedding impacts on a block level rather than a pixel level. Then, a policy network is utilized to learn optimal joint embedding policies for pixel blocks through interactions with the environment. Afterwards, these policies can be converted into joint embedding costs for practical message embedding. The structure of the policy network is designed with an effective attention mechanism and incorporated with the domain knowledge derived from traditional non-additive steganographic methods. The environment is responsible for assigning rewards according to the impacts of the sampled joint embedding actions, which are evaluated by the gradient information of a neural network-based steganalyzer. Experimental results show that the proposed non-additive method JoPoL significantly outperforms the existing additive methods against both feature-based and CNN-based steganalzyers over different payloads. [ABSTRACT FROM AUTHOR]

Though there are many encouraging reports, existing image colorization algorithms are still prone to unnatural visual distortions. We observe that unnatural visual distortions are mainly introduced in the deconvolutional modules of existing generative models. Furthermore, the existing algorithms are with heavily structures, which hinders the deployment of algorithms on edge devices. In this paper, we propose ISP-GAN, a novel lightweight generative adversarial network with inception sub-pixel deconvolution aimed at improving the performance of image colorization. In the generator of our proposed ISP-GAN, we propose a novel inception sub-pixel deconvolutional block (ISP), along with a modified residual convolutional block (MRC), to avoid representational bottlenecks and consequently expand perceptual fields. For the ISP-GAN discriminator, we apply deep-learning-based steganalytic networks to improve the training efficiency of the whole framework and consequently enhance the performance of the corresponding generator. Our ISP-GAN is with lightweight structures and experimental results on the benchmark datasets show that ISP-GAN can achieve state-of-the-art performance in the image colorization task. [ABSTRACT FROM AUTHOR]

The Jominy end-quench (JEQ) test was adopted to investigate the quench sensitivity of Al-6.6Zn-1.8Mg-0.23Cu-0.22Mn-0.21Zr (7046A) alloy. The mechanical properties of the aged alloy were studied by means of a hardness and tensile property test, whereas microstructure observation was realized by x-ray diffraction (XRD) and transmission electron microscopy (TEM). The results show the following: (I) inhomogeneity exists in the mechanical properties of as-aged alloy and they generally decrease with the distance from the quenched end of the bar (D) where the depth of losing 10% of maximum hardness is 196 mm. (II) Average cooling rates decreased with the increase of D. When the critical average cooling rate is about 1 K s−1, 90% of the maximum hardness can be obtained. (III) On the other hand, the influence of Al3Zr dispersoids and Al6Mn dispersoids on the precipitate behavior of MgZn2 was studied, indicating that Al3Zr and Al6Mn were effective heterogeneous precipitates sites in the slow quenching process. [ABSTRACT FROM AUTHOR]

The mechanical properties, corrosion behavior and microstructures of the Al-Zn-Mg-Cu alloy under various ageing treatments were investigated comparatively. The results show that the tensile strength and corrosion resistance are strongly affected by the precipitate state. Massive fine intragranular precipitates contribute to high strength. Discontinuous coarse grain boundary precipitates containing high Cu content, as well as the narrow precipitate free zone, result in low corrosion susceptibility. After the non-isothermal ageing (NIA) treatment, the tensile strength of 577 MPa is equivalent to that of 579 MPa for the T6 temper. Meanwhile, the stress corrosion susceptibility rtf and the maximum corrosion depth are 97.8% and 23.5 μm, which are comparable to those of 92.8% and 26.7 μm for the T73 temper. Moreover, the total ageing time of the NIA treatment is only 7.25 h, which is much less than that of 48.67 h for the retrogression and re-ageing condition. [ABSTRACT FROM AUTHOR]

Median filtering, being an order statistic filtering, has been widely used in image denoising and recently also in image anti-forensics and anti-steganalysis. In the past few years, several methods have been developed for median filtering detection. However, it is still a challenging task to detect median filtering in JPEG compressed images. In this paper, we propose a novel method to solve this challenging task. We first generate median filtered residual (MFR), average filtered residual (AFR) and Gaussian filtered residual (GFR) by calculating the differences between an original image and its filtered images. Then, we propose to use two-dimensional autoregressive (2D-AR) model to characterize MFR, AFR and GFR separately, and further combine the 2D-AR coefficients of these three residuals into a set of features. Finally, the extracted feature set is fed into a support vector machine classifier for training and detection. Extensive experiments have demonstrated that compared with existing methods, the proposed one can achieve a considerable improvement in detecting median filtering in heavily compressed images. [ABSTRACT FROM AUTHOR]

The tubers of Cremastra appendiculata (D. Don) Makino (Orchidaceae) are used in traditional Chinese medicine for the treatment of cancer and bacterial infections. Its 95% ethanol extract was found to display potent inhibitory activities on butyrylcholinesterase (BChE) (IC50 = 23.66 µg/mL) and β-amyloid peptide aggregation (74.09% at 100 μg/mL). Active ethyl acetate extract afforded one bibenzyl (1), four phenanthrenes (2-5), and one rotenoid (6) by a bioassay-guided isolation. Compounds 2, 3, and 5 exhibited potent BChE inhibitory effects with IC50 values of 19.66, 32.80, and 37.79 µM, respectively. Kinetic studies indicated that both 2 and 3 were mixed-type BChE inhibitors. Their strong binding affinities to BChE were further confirmed by fluorescence quenching analysis. Meanwhile, 2 and 3 showed the effective DPPH and ABTS radical scavenging activities (EC50 < 11 μM) and inhibition effect on β-amyloid peptide aggregation (64.49 and 29.50% at 20 μM, respectively), suggesting they could serve as multifunctional potential agents for AD drugs development. This paper revealed the potential medicinal and economic values of C. appendiculata and its chemical constituents in AD prevention/treatment. [ABSTRACT FROM AUTHOR]

Based on digital watermark, a speech forensics scheme is proposed. The feature coefficients cross-correlation degree of speech signal is defined, and the property is discussed, which demonstrates that the feature is very robust. Then a new watermark embedding method based on the feature is explored, aiming to enlarge the embedding capacity and solve the security issue of watermark schemes based on public features. In this paper, for each fame of speech signal, it is cut into two parts, and each part is divided into some segments. Then frame number is mapped to a sequence of integers, which are embedded into the segments. The integers can be extracted used for forensics and tamper location after watermarked signal being attacked. Theoretical analysis and experimental results show that the scheme proposed is inaudible and robust against desynchronization attacks, enhances the security of watermark system and has a good ability for speech forensics. [ABSTRACT FROM AUTHOR]

A secure multiparty quantum key agreement protocol using single-qubit states is proposed. The agreement key is computed by performing exclusive-OR operation on all the participants' secret keys. Based on the commutative property of the commutative encryption, the exclusive-OR operation can be performed on the plaintext in the encrypted state without decrypting it. Thus, it not only protects the final shared key, but also reduces the complexity of the computation. The efficiency of the proposed protocol, compared with previous multiparty QKA protocols, is also improved. In the presented protocol, entanglement states, joint measurement and even the unitary operations are not needed, and only rotation operations and single-state measurement are required, which are easier to be realized with current technology. [ABSTRACT FROM AUTHOR]

Double JPEG compression detection plays an important role in digital image forensics. Recently, Huang et al. (IEEE Trans Inf Forensics Security 5(4):848-856, 2010) first pointed out that the number of different discrete cosine transform (DCT) coefficients would monotonically decrease when repeatedly compressing a JPEG image with the same quantization matrix, and a strategy based on random permutation was developed to expose such an operation successfully. In this paper, we propose an anti-forensic method to fool this method. The proposed method tries to slightly modify the DCT coefficients for confusing the traces introduced by double JPEG compression with the same quantization matrix. By investigating the relationship between the DCT coefficients of the first compression and those of the second one, we determine the quantity of modification by constructing a linear model. Furthermore, in order to improve the security of anti-forensics, the locations of modification are adaptively selected according to the complexity of the image texture. The extensive experiments evaluated on 10,000 natural images have shown that the proposed method can effectively confuse the detector proposed in Huang et al. (IEEE Trans Inf Forensics Security 5(4):848-856, 2010), while keeping higher visual quality and leaving fewer other detectable statistical artifacts. [ABSTRACT FROM AUTHOR]

Video frame-rate up-conversion is one of the common operations for tampering digital videos in the temporal domain, such as creating fake high-quality videos and splicing two video clips with different frame rates. However, few existing works have been proposed for detecting this form of tampering operation. Based on the analysis of extensive experiments, we found that frame-rate up-conversion algorithms employed in most current video editing softwares will inevitably introduce some periodic artifacts into inter-frame similarity in the resulting video frame sequence. By analyzing such artifacts, we propose a simple yet very effective method to expose video after frame-rate up-conversion, and further estimate its original frame rate. The experimental results evaluated on 100 original videos at different frame rates have shown the effectiveness of the proposed method. The average detection accuracy can achieve as high as 99 % on noise-free videos in uncompressed and H.264/AVC formats. Besides, the proposed method is robust to noise as the detection accuracy could reach over 85 % and 95 % on noised videos with Gaussian white noise when SNR is 33 db and 36 db respectively. [ABSTRACT FROM AUTHOR]

Non-intrusive digital image forensics (NIDIF) is a novel approach to authenticate the trustworthiness of digital images. It works by exploring varieties of intrinsic characteristics involved in the digital imaging, editing, storing processes as discriminative features to reveal the subtle traces left by a malicious fraudster. The NIDIF for the lossy JPEG image format is of special importance for its pervasive application. In this paper, we propose an NIDIF framework for the JPEG images. The framework involves two complementary identification methods for exposing shifted double JPEG (SD-JPEG) compression artifacts, including an improved ICA-based method and a First Digits Histogram based method. They are designed to treat the detectable conditions and a few special undetectable conditions separately. Detailed theoretical justifications are provided to reveal the relationship between the detectability of the artifacts and some intrinsic statistical characteristics of natural image signal. The extensive experimental results have shown the effectiveness of the proposed methods. Furthermore, some case studies are also given to demonstrate how to reveal certain types of image manipulations, such as cropping, splicing, or both, with our framework. [ABSTRACT FROM AUTHOR]

In this paper, we propose a robust perceptual hashing algorithm by using video luminance histogram in shape. The underlying robustness principles are based on three main aspects: 1) Since the histogram is independent of position of a pixel, the algorithm is resistant to geometric deformations; 2) the hash is extracted from the spatial Gaussian-filtering low-frequency component for those common video processing operations such as noise corruption, low-pass filtering, lossy compression, etc.; 3) a temporal Gaussian-filtering operation is designed so that the hash is resistant to temporal desynchronization operations, such as frame rate change and dropping. As a result, the hash function is robust to common geometric distortions and video processing operations. Experimental results show that the proposed hashing strategy can provide satisfactory robustness and uniqueness. [ABSTRACT FROM AUTHOR]

Rendering technology in computer graphics (CG) is now capable of producing highly photorealistic images, giving rise to the problem of how to identify CG images from natural images. Some methods were proposed to solve this problem. In this paper, we give a novel method from a new point of view of image perception. Although the photorealistic CG images are very similar to natural images, they are surrealistic and smoother than natural images, thus leading to the difference in perception. A part of features are derived from fractal dimension to capture the difference in color perception between CG images and natural images, and several generalized dimensions are used as the rest features to capture difference in coarseness. The effect of these features is verified by experiments. The average accuracy is over 91.2%. [ABSTRACT FROM AUTHOR]

Resistance to ambiguity attack is an important requirement for a secure digital rights management (DRM) system. In this paper, we revisit the non-ambiguity of a blind watermarking based on the computational indistinguishability between pseudo random sequence generator (PRSG) sequence ensemble and truly random sequence ensemble. Ambiguity attacker on a watermarking scheme, which uses a PRSG sequence as watermark, is viewed as an attacker who tries to attack a noisy PRSG sequence. We propose and prove the security theorem for binary noisy PRSG sequence and security theorem for general noisy PRSG sequence. It is shown that with the proper choice of the detection threshold Th = a √n ( a is a normalized detection threshold; n is the length of a PRSG sequence) and n ⩾ 1.39 × m/ a 2 ( m is the key length), the success probability of an ambiguity attack and the missed detection probability can both be made negligibly small thus non-ambiguity and robustness can be achieved simultaneously for both practical quantization-based and blind spread spectrum (SS) watermarking schemes. These analytical resolutions may be used in designing practical non-invertible watermarking schemes and measuring the non-ambiguity of the schemes. [ABSTRACT FROM AUTHOR]

For steganalysis of JPEG images, features derived in the embedding domain appear to achieve a preferable performance. However, with the existing JPEG steganography, the minor changes due to the hidden secret data are not easy to be explored directly from the quantized block DCT (BDCT) coefficients in that the energy of the carrier image is much larger than that of the hidden signal. In this paper, we present an improved calibration-based universal JPEG steganalysis, where the microscopic and macroscopic calibrations are combined to calibrate the local and global distribution of the quantized BDCT coefficients of the test image. All features in our method are generated from the difference signal between the quantized BDCT coefficients of the test image and its corresponding microscopic calibrated image, or calculated as the difference between the signal extracted from test image and its corresponding macroscopic calibrated image. The extracted features will be more effective for our classification. Moreover, through using the Markov empirical transition matrices, both magnitude and sign dependencies along row scanning and column scanning patterns existed in intra-block and inter-block quantized BDCT coefficients are employed in our method. Experimental results demonstrate that our proposed scheme outperforms the best effective JPEG steganalyzers having been presented. [ABSTRACT FROM AUTHOR]

The Ate pairing has been suggested since it can be computed efficiently on ordinary elliptic curves with small values of the traces of Frobenius t. However, not all pairing-friendly elliptic curves have this property. In this paper, we generalize the Ate pairing and find a series of the variations of the Ate pairing. We show that the shortest Miller loop of the variations of the Ate pairing can possibly be as small as r 1/ φ( k) on some special pairing-friendly curves with large values of Frobenius trace, and hence speed up the pairing computation significantly. [ABSTRACT FROM AUTHOR]

Pairing-based cryptosystems have developed very fast in the last few years. The efficiencies of these cryptosystems depend on the computation of the bilinear pairings. In this paper, a new efficient algorithm based on double-base chains for computing the Tate pairing is proposed for odd characteristic p > 3. The inherent sparseness of double-base number system reduces the computational cost for computing the Tate pairing evidently. The new algorithm is 9% faster than the previous fastest method for the embedding degree k = 6. [ABSTRACT FROM AUTHOR]

The 21st century is the age of information when information becomes an important strategic resource. The information obtaining, processing and security guarantee capability are playing critical roles in comprehensive national power, and information security is related to the national security and social stability. Therefore, we should take measures to ensure the information security of our country. In recent years, momentous accomplishments have been obtained with the rapid development of information security technology. There are extensive theories about information security and technology. However, due to the limitation of length, this article mainly focuses on the research and development of cryptology, trusted computing, security of network, and information hiding, etc. [ABSTRACT FROM AUTHOR]

Over the past years, digital images have been widely used in the Internet and other applications. Whilst image processing techniques are developing at a rapid speed, tampering with digital images without leaving any obvious traces becomes easier and easier. This may give rise to some problems such as image authentication. A new passive technology for image forensics has evolved quickly during the last few years. Unlike the signature-based or watermark-based methods, the new technology does not need any signature generated or watermark embedded in advance. It assumes that different imaging devices or processing would introduce different inherent patterns into the output images. These underlying patterns are consistent in the original untampered images and would be altered after some kind of manipulations. Thus, they can be used as evidence for image source identification and alteration detection. In this paper, we will discuss this new forensics technology and give an overview of the prior literatures. Some concluding remarks are made about the state of the art and the challenges in this novel technology. [ABSTRACT FROM AUTHOR]

Stochastic modulation steganography embeds secret message within the cover image by adding stego-noise with a specific probabilistic distribution. No method is known to be applicable to the estimation of stochastic modulation steganography. By analyzing the distributions of the horizontal pixel difference of images before and after stochastic modulation embedding, we present a new steganalytic approach to accurately estimate the length of secret message in stochastic modulation steganography. The proposed method first establishes a model describing the statistical relationship among the differences of the cover image, stego-image and stego-noise. In the case of stego-image-only steganalysis, rough estimate of the distributional parameters of the cover image’s pixel difference is obtained with the use of the provided stego-image. And grid search and Chi-square goodness of fit test are exploited to estimate the length of the secret message embedded with stochastic modulation steganography. The experimental results demonstrate that our new approach is effective for steganalyzing stochastic modulation steganography and accurately estimating the length of the secret message. [ABSTRACT FROM AUTHOR]

Nowadays, the abuse of deepfakes is a well-known issue since deepfakes can lead to severe security and privacy problems. And this situation is getting worse, as attackers are no longer limited to unimodal deepfakes, but use multimodal deepfakes, i.e., both audio forgery and video forgery, to better achieve malicious purposes. The existing unimodal or ensemble deepfake detectors are demanded with fine-grained classification capabilities for the growing technique on multimodal deepfakes. To address this gap, we propose a graph attention network based on heterogeneous graph for fine-grained multimodal deepfake classification, i.e., not only distinguishing the authenticity of samples, but also identifying the forged types, e.g., video or audio or both. To this end, we propose a positional coding-based heterogeneous graph construction method that converts an audio-visual sample into a multimodal heterogeneous graph according to relevant hyperparameters. Moreover, a cross-modal graph interaction module is devised to utilize audio-visual synchronization patterns for capturing inter-modal complementary information. The de-homogenization graph pooling operation is elaborately designed to keep differences in graph node features for enhancing the representation of graph-level features. Through the heterogeneous graph attention network, we can efficiently model intra- and inter-modal relationships of multimodal data both at spatial and temporal scales. Extensive experimental results on two audio-visual datasets FakeAVCeleb and LAV-DF demonstrate that our proposed model obtains significant performance gains as compared to other state-of-the-art competitors. The code is available at https://github.com/yinql1995/Fine-grained-Multimodal-DeepFake-Classification/. [ABSTRACT FROM AUTHOR]