Showing papers on "Watermark published in 2020"

Adversarial Frontier Stitching for Remote Neural Network Watermarking

Abstract: The state-of-the-art performance of deep learning models comes at a high cost for companies and institutions, due to the tedious data collection and the heavy processing requirements. Recently, Nagai et al. (Int J Multimed Inf Retr 7(1):3–16, 2018), Uchida et al. (Embedding watermarks into deep neural networks, ICMR, 2017) proposed to watermark convolutional neural networks for image classification, by embedding information into their weights. While this is a clear progress toward model protection, this technique solely allows for extracting the watermark from a network that one accesses locally and entirely. Instead, we aim at allowing the extraction of the watermark from a neural network (or any other machine learning model) that is operated remotely, and available through a service API. To this end, we propose to mark the model’s action itself, tweaking slightly its decision frontiers so that a set of specific queries convey the desired information. In the present paper, we formally introduce the problem and propose a novel zero-bit watermarking algorithm that makes use of adversarial model examples. While limiting the loss of performance of the protected model, this algorithm allows subsequent extraction of the watermark using only few queries. We experimented the approach on three neural networks designed for image classification, in the context of MNIST digit recognition task.

Deep Reinforcement Learning for Resource Protection and Real-Time Detection in IoT Environment

Abstract: With the fast advancements of electronic chip technologies in the Internet of Things (IoT), it is urgent to address the copyright protection issue of intellectual property (IP) circuit resources of the electronic devices in IoT environments. In this article, a fast deep-reinforcement-learning (DRL)-based detection algorithm for virtual IP watermarks is proposed by combining the technologies of mapping function and DRL to preprocess the ownership information of the IP circuit resource. The deep $Q$ -learning (DQN) algorithm is used to generate the watermarked positions adaptively, making the watermarked positions secure yet close to the original design, turning the watermarked positions secure. An artificial neural network (ANN) algorithm is utilized for training the position distance characteristic vectors of the IP circuit, in which the characteristic function of the virtual position for IP watermark is generated after training. In IP ownership verification, the DRL model can quickly locate the range of virtual watermark positions. With the characteristic values of the virtual positions in each lookup table (LUT) area and surrounding areas, the mapping position relationship can be calculated in a supervised manner in the neural network, as the algorithm realizes the fast location of the real ownership information in an IP circuit. The experimental results show that the proposed algorithm can effectively improve the speed of watermark detection as also reducing the resource overhead. Besides, it also achieves excellent performance in security.

ReDMark: Framework for Residual Diffusion Watermarking based on Deep Networks

Abstract: Due to the rapid growth of machine learning tools and specifically deep networks in various computer vision and image processing areas, applications of Convolutional Neural Networks for watermarking have recently emerged. In this paper, we propose a deep end-to-end diffusion watermarking framework (ReDMark) which can learn a new watermarking algorithm in any desired transform space. The framework is composed of two Fully Convolutional Neural Networks with residual structure which handle embedding and extraction operations in real-time. The whole deep network is trained end-to-end to conduct a blind secure watermarking. The proposed framework simulates various attacks as a differentiable network layer to facilitate end-to-end training. The watermark data is diffused in a relatively wide area of the image to enhance security and robustness of the algorithm. Comparative results versus recent state-of-the-art researches highlight the superiority of the proposed framework in terms of imperceptibility, robustness and speed. The source codes of the proposed framework are publicly available at Github 1 .

KeySplitWatermark: Zero Watermarking Algorithm for Software Protection Against Cyber-Attacks

Abstract: Cyber-attacks are evolving at a disturbing rate. Data breaches, ransomware attacks, crypto-jacking, malware and phishing attacks are now rampant. In this era of cyber warfare, the software industry is also growing with an increasing number of software being used in all domains of life. This evolution has added to the problems of software vendors and users where they have to prevent a wide range of attacks. Existing watermark detection solutions have a low detection rate in the software. In order to address this issue, this paper proposes a novel blind Zero code based Watermark detection approach named KeySplitWatermark, for the protection of software against cyber-attacks. The algorithm adds watermark logically into the code utilizing the inherent properties of code and gives a robust solution. The embedding algorithm uses keywords to make segments of the code to produce a key-dependent on the watermark. The extraction algorithms use this key to remove watermark and detect tampering. When tampering increases to a certain user-defined threshold, the original software code is restored making it resilient against attacks. KeySplitWatermark is evaluated on tampering attacks on three unique samples with two distinct watermarks. The outcomes show that the proposed approach reports promising results against cyber-attacks that are powerful and viable. We compared the performance of our proposal with state-of-the-art works using two different software codes. Our results depict that KeySplitWatermark correctly detects watermarks, resulting in up to 15.95 and 17.43 percent reduction in execution time on given code samples with no increase in program size and independent of watermark size.

Chaotic based secure watermarking approach for medical images

Abstract: In this paper, a chaotic based secure medical image watermarking approach is proposed. The method is using non sub-sampled contourlet transform (NSCT), redundant discrete wavelet transform (RDWT) and singular value decomposition (SVD) to provide significant improvement in imperceptibility and robustness. Further, security of the approach is ensured by applying 2-D logistic map based chaotic encryption on watermarked medical image. In our approach, the cover image is initially divided into sub-images and NSCT is applied on the sub-image having maximum entropy. Subsequently, RDWT is applied to NSCT image and the singular vector of the RDWT coefficient is calculated. Similar procedure is followed for both watermark images. The singular value of both watermarks is embedded into the singular matrix of the cover. Experimental evaluation shows when the approach is subjected to attacks, using combination of NSCT, RDWT, SVD and chaotic encryption it makes the approach robust, imperceptible, secure and suitable for medical applications.

A block-based RDWT-SVD image watermarking method using human visual system characteristics

Abstract: With the rapid growth of internet technology, image watermarking method has become a popular copyright protection method for digital images. In this paper, we propose a watermarking method based on $$4\times 4$$ image blocks using redundant wavelet transform with singular value decomposition considering human visual system (HVS) characteristics expressed by entropy values. The blocks which have the lower HVS entropies are selected for embedding the watermark. The watermark is embedded by examining $$U_{2,1}$$ and $$U_{3,1}$$ components of the orthogonal matrix obtained from singular value decomposition of the redundant wavelet transformed image block where an optimal threshold value based on the trade-off between robustness and imperceptibility is used. In order to provide additional security, a binary watermark is scrambled by Arnold transform before the watermark is embedded into the host image. The proposed scheme is tested under various image processing, compression and geometrical attacks. The test results are compared to other watermarking schemes that use SVD techniques. The experimental results demonstrate that our method can achieve higher imperceptibility and robustness under different types of attacks compared to existing schemes. Our method provides high robustness especially under image processing attacks, JPEG2000 and JPEG XR attacks. It has been observed that the proposed method achieves better performance over the recent existing watermarking schemes.

Secure data hiding techniques: a survey

Abstract: This article presents a detailed discussion of different prospects of digital image watermarking. This discussion of watermarking included: brief comparison of similar information security techniques, concept of watermark embedding and extraction process, watermark characteristics and applications, common types of watermarking techniques, major classification of watermarking attacks, brief summary of various secure watermarking techniques. Further, potential issues and some existing solutions are provided. Furthermore, the performance comparisons of the discussed techniques are presented in tabular format. Authors believe that this article contribution will provide as catalyst for potential researchers to implement efficient watermarking systems.

Color image watermarking based on orientation diversity and color complexity

Abstract: The Just Noticeable Distortion (JND) can reliably measure the perceptual strength in image watermarking, but, it remains a challenge to computationally model the process of embedding watermark without prior knowledge of the image contents. This paper proposed a novel color image watermarking scheme in discrete cosine transform (DCT) domain based on JND, which takes both orientation diversity and color complexity features into account. Firstly, two indicator was introduced which take into account the differences in the texture types and orientation diversity of the Human Visual System (HVS) in the proposed JND contrast masking (CM) processing. In addition, a novel color complexity weight from Cb-channel is used to guarantee the scheme robustness. Then, a novel JND model combined with the proposed contrast masking and color complexity is applied into quantization watermarking scheme. Compared with the state-of-the-art methods for color image watermarking, experimental results using publicly available images show that our proposed scheme is reliable and effective.

Hybrid Watermarking Algorithm Using Clifford Algebra With Arnold Scrambling and Chaotic Encryption

Abstract: With the widespread use of color images, the copyright protection of those images using watermarks is one of the latest research topics. The use of color images as watermarks has advantages over binary and irreplaceable grayscale images. Color images are intuitive, rich, and lively; they have large amounts of copyright protection information and more easily recognized by human vision. To improve the security of watermark information and embedding positions and improve the algorithm’s robustness against various attacks, a Quaternion Fourier transform (QFT) based algorithm, based on Arnold transform and chaotic encryption, is proposed in this paper. Geometric algebra (GA) can deal with color images in vector form with each component of RGB handled individually. We used Quaternion, which is a sub-algebra of GA, and effectively handled color image processing by using Fourier transformation. After deriving the calculation process of the QFT with strong security by Arnold scrambling and chaotic encryption, this paper proposes a digital watermarking algorithm that resists geometric attacks by using color images as carriers. The robustness and quality of the proposed watermarking algorithm is tested with different with many statistical measures. Experimental outcomes show that the proposed approach is the best to solve conflict problems between quality and robustness. Also, the proposed approach exhibits worthy robustness against many attacks, such as, conventional attacks, and geometrical attacks.

A data authentication scheme for UAV ad hoc network communication

Abstract: In this paper, we propose an efficient and energy-saving distributed network architecture based on clustering stratification to solve the information security problem of unmanned aerial vehicle ad hoc network communication. And a double-authentication watermarking strategy is designed. In order to ensure that the data collected by nodes can be sent securely to the cluster head node, we use the self-characteristic of the collected data to generate the authentication watermark and insert it into the collected data at random. The cluster head node first verifies the integrity of collected data and deletes the suspicious data. Then, the authentication information is generated by combining the chaotic mapping method, and the watermark is hidden by changing the parity of the least significant bits of the data. Experimental results show that the proposed security strategy can resist most attacks, such as selective forwarding, data replay and tampering. Meanwhile, it has low energy consumption and low latency.

A New Chaotic Image Watermarking Scheme Based on SVD and IWT

Abstract: Image watermarking schemes based on singular value decomposition (SVD) have become popular due to a good trade-off between robustness and imperceptibility. However, the false positive problem (FPP) is the main drawback of SVD-based watermarking schemes. The singular value is the main cause of FPP issues because it a fixed value that does not hold structural information of an image. In this paper, a new SVD-based image watermarking scheme that uses a chaotic map is proposed to overcome this issue. The secret key is first extracted from both the host and watermark image. This key is used to generate a new chaotic matrix and chaotic multiple scaling factors (CMSF) to increase the sensitivity of the proposed scheme. The watermark image is then transformed based on the chaotic matrix before being directly embedded into the singular value of the host image by using the CMSF. The extracted secret key is unique to the host and the watermark images, which improves security and overcomes FPP issues. Experimental results show that the proposed scheme fulfils all watermarking requirements in terms of robustness, imperceptibility, security, and payload. Furthermore, it achieves high robustness with different scaling factors, and outperforms several existing schemes.

SVM-based robust image watermarking technique in LWT domain using different sub-bands

Abstract: In this paper, a robust image watermarking system in lifting wavelet transform domain using different sub-bands has been proposed. SVM classifier is used during watermark extraction to obtain improved robustness under diverse attack conditions. In this work, a detailed analysis of imperceptibility and robustness performance with the use of different sub-bands has been presented. The performance on different sub-band has been analyzed so as to maximize the robustness against different attacks keeping imperceptibility at adequate level. Robustness is observed against various attacks such as noising attacks, denoising attacks, image processing attacks, lossy compression attacks and geometric attacks. It is seen that high-frequency sub-band provides better invisibility, whereas variation of robustness performance on different sub-bands depend on the type of attacks. It is observed from the performance analysis that all the attacks do not have exactly same effect on the frequency content of the image. For instance, noising attack affects every frequency component of the image almost equally, whereas the embedding in high-frequency band makes the system fragile to lossy compression attack. The algorithm is tested on a large image database to observe the variation in the performance of the system. Comparative analysis suggests that the proposed sub-band provides improved performance over some benchmark methods in most of the cases.

Model Watermarking for Image Processing Networks.

Abstract: Deep learning has achieved tremendous success in numerous industrial applications. As training a good model often needs massive high-quality data and computation resources, the learned models often have significant business values. However, these valuable deep models are exposed to a huge risk of infringements. For example, if the attacker has the full information of one target model including the network structure and weights, the model can be easily finetuned on new datasets. Even if the attacker can only access the output of the target model, he/she can still train another similar surrogate model by generating a large scale of input-output training pairs. How to protect the intellectual property of deep models is a very important but seriously under-researched problem. There are a few recent attempts at classification network protection only.In this paper, we propose the first model watermarking framework for protecting image processing models. To achieve this goal, we leverage the spatial invisible watermarking mechanism. Specifically, given a black-box target model, a unified and invisible watermark is hidden into its outputs, which can be regarded as a special task-agnostic barrier. In this way, when the attacker trains one surrogate model by using the input-output pairs of the target model, the hidden watermark will be learned and extracted afterward. To enable watermarks from binary bits to high-resolution images, both traditional and deep spatial invisible watermarking mechanism are considered. Experiments demonstrate the robustness of the proposed watermarking mechanism, which can resist surrogate models learned with different network structures and objective functions. Besides deep models, the proposed method is also easy to be extended to protect data and traditional image processing algorithms.

Robust and hybrid SVD-based image watermarking schemes:

Abstract: These days, researchers have used many techniques in e-multimedia data for intellectual property protection. One of these important techniques is watermarking which is superior to Digital Signature because it does not rise overhead and it can prevent unauthorized users from misusing digital information. However, the number of attacks with aim of distorting watermark signal or disturbing watermarks detecting are exploding. Consequently, the watermark techniques have became more complicated, but still they are not fully robust against some geometric distortions. To overcome this issue and to be enough robust against the malicious attacks, and provide a high level of imperceptibility,the watermark schemes have discovered and developed the advantages by means of Singular-value decomposition. Therefore, in recent decades there has been a significant increasing in using of SVD domain in many new image watermarking schemes, which show high robustness and imperceptibility. This discussion of SVD in image watermarking schemes included: a concise overview on watermarking systems, a brief review on SVD properties in image processing and its weaknesses, the concept of SVD components, the main classification of different SVD based image-watermarking schemes; a brief summary of various secure SVD-based watermarking techniques and all major SVD coefficients modifying strategies are presented. In addition, potential issues and some existing solutions and guidelines are provided. Furthermore, the performance comparisons of the discussed techniques are presented in tabular format. This article contribution provides essential and potential information for the researchers to design, develop and implement efficient robust and hybrid SVD-based image watermarking systems.

An improved watermarking algorithm for color image using Schur decomposition

Abstract: In order to protect the color image copyright protection of the multimedia big data, it is necessary to design a color image watermarking algorithm. To achieve this purpose, an improved color image watermarking algorithm based on Schur decomposition is proposed in this paper. First, the watermark information is, respectively, embedded into the upper triangular matrix and the unitary matrix of Schur decomposition by two different methods, and two temporary watermarked image blocks are obtained. Then, the proposed improved method is used to select the final watermarked image block from these temporary watermarked image blocks. The highlight of the proposed method is that the final watermarked block has less visual distortion. Meanwhile, the embedded flag is created and uploaded to the cloud service provider with the watermarked image. When extracting watermark, the original host image or the watermark image is not needed. Experimental results show that the proposed watermarking algorithm has better performance; in particular, the watermark invisibility has been obviously improved than other methods considered in this paper.

Blind watermarking for color images using EMMQ based on QDFT

Abstract: In this study, we developed a novel scheme for the blind watermarking of color images. The proposed scheme incorporates extreme pixel adjustment (EPA), multi-bit partly sign-altered mean modulation (MPSAM), mixed modulation (MM), and particle swarm optimization (PSO) within a scheme based on crisscross inter-block quaternion discrete Fourier transform (QDFT). Accordingly, the proposed scheme employing EPA, MPSAM, MM, and QDFT is referred to as EMMQ. The image is separated into non-overlapping 8 × 8 pixel blocks, whereupon MPSAM is used to map multiple bits within a single block using multiple coefficients in one of the four transformed components of QDFT. The use of MM to embed the watermark allows for superior image quality and strong resistance to image processing attacks. Our use of PSO also makes it possible to optimize the EMMQ parameters, thereby enabling outstanding robustness without compromising imperceptibility. Experiment results demonstrate the efficacy of the proposed scheme in resisting a variety of image-processing attacks, with performance superior to that of existing watermarking schemes in terms of imperceptibility and robustness for a given payload capacity.

A blind and robust color images watermarking method based on block transform and secured by modified 3-dimensional Hénon map

Abstract: This paper proposes a blind and robust color image watermarking method based on a new three-dimensional Henon chaotic map and uses integer wavelet transform, discrete wavelet transform and contourlet transform in embedding and extracting processes. In the presented approach, color images are divided into $$4\times 4$$ main nonoverlapping parts, and one of the transforms is applied to these parts. Then the low–low sub-band of transform is selected. The suggested map is used to find $$2\times 2$$ blocks in the embedding process. The bits of watermark are embedded in the parts of images to increase the robustness of the proposed watermarking scheme. To improve the quality of the final watermark, the suggested technique uses a correction process in the extracting process. In this paper, the bifurcation diagram, Lyapunov exponent, cobweb plot and trajectory diagram are used to show the chaotic behavior of the proposed map. Based on DIEHARD, ENT and NIST test suites, the suggested map can be used as a pseudo-random number generator. The simulation results show that the proposed watermarking algorithm is robust against most image processing attacks like salt & pepper, cropping, low-pass filter, wiener filter, blurring, etc. The comparison results between the suggested watermarking scheme, and some similar methods show that the presented technique has good performance, imperceptibility, acceptable robustness and outperforms most related methods.

Efficient SVD-based audio watermarking technique in FRT domain

Abstract: This paper presents an audio watermarking technique based on singular value decomposition (SVD) and fractional Fourier transform (FRT). The basic idea of this technique is to implement SVD watermarking on the audio signals in the FRT domain due to its recommended degree of security resulting from using a rotation angle in addition to the frequency-domain transformation. The SVD has an invariance to changes in the signal after watermark embedding. Hence, the proposed technique has a large degree of security and resistance to attacks. This technique is based on embedding an image watermark in either the audio signal or a transformed version of this signal. Experimental results show that watermark embedding in the FRT of an audio signal achieves less distortion of the audio signal in the absence of attacks. In the presence of attacks, it is recommended that the embedding is performed in the FRT of the audio signal to maintain a high detection correlation coefficient between the original watermark and the obtained watermark. A segment-based implementation of the proposed audio watermarking technique is also presented. This implementation succeeds in obtaining a high detection correlation coefficient in the presence of severe attacks. It is noticed from the results that in the presence of attacks, the SVD watermarking in the FRT domain with a phase angle of 5π/4 is better for watermark detection than watermarking using other angles in the FRT domain.

Multi-dimensional particle swarm optimization for robust blind image watermarking using intertwining logistic map and hybrid domain

Abstract: Robust blind watermarking has become a vital means of copyright protection, and this paper presents a new optimal robust and blind watermarking method of grayscale images based on intertwining logistic map and a variant of particle swarm optimization (PSO) in a hybrid domain. In the proposed approach, firstly a host image is decomposed by discrete wavelet transform, and discrete cosine transform (DCT) is applied to insensitive LH and HL subbands according to human visual model. Then, optimum frequency spectra in the DCT domain are chosen to form a feature matrix for improving the robustness and transparency of watermark. Finally, a shuffled watermark image using the chaotic logistic map is inserted by modifying the largest singular values of a feature matrix pair in the singular value decomposition domain. An improved version of PSO is employed to perform multi-dimensional optimization for selection of the most qualified DCT coefficients and estimation of watermark embedding strength in terms of their significant influence on imperceptibility and robustness. The security of the proposed method is provided by intertwining logistic map. Experimental results on standard test images indicate that PSO searches efficiently optimal values of watermark embedding strength and the most suitable DCT subbands, and the proposed watermarking algorithm performs much better than the other compared schemes in imperceptibility and robustness objectives.

A Pixel-Based Digital Medical Images Protection Using Genetic Algorithm with LSB Watermark Technique

Abstract: In recent years, especially when these images are transmitted via a network, digital image security has attracted attention. The 2D barcode is designed to encrypt patient information. In this paper, we proposed a grayscale medical image encryption technology based on the characteristics of genetic algorithm (GAS) using LSB technology. The patient information will be converting into 2D barcode and after that, the original image is embedded with 2D barcode using LSB technique. The LSB technique is more suitable for medical imaging. The resulting image is subjected to a genetic algorithm. In this purpose design, we combined the genetic with LSB technique to make the patient image and its information more secured.

Model Watermarking for Image Processing Networks

Abstract: Deep learning has achieved tremendous success in numerous industrial applications. As training a good model often needs massive high-quality data and computation resources, the learned models often have significant business values. However, these valuable deep models are exposed to a huge risk of infringements. For example, if the attacker has the full information of one target model including the network structure and weights, the model can be easily finetuned on new datasets. Even if the attacker can only access the output of the target model, he/she can still train another similar surrogate model by generating a large scale of input-output training pairs. How to protect the intellectual property of deep models is a very important but seriously under-researched problem. There are a few recent attempts at classification network protection only. In this paper, we propose the first model watermarking framework for protecting image processing models. To achieve this goal, we leverage the spatial invisible watermarking mechanism. Specifically, given a black-box target model, a unified and invisible watermark is hidden into its outputs, which can be regarded as a special task-agnostic barrier. In this way, when the attacker trains one surrogate model by using the input-output pairs of the target model, the hidden watermark will be learned and extracted afterward. To enable watermarks from binary bits to high-resolution images, both traditional and deep spatial invisible watermarking mechanism are considered. Experiments demonstrate the robustness of the proposed watermarking mechanism, which can resist surrogate models learned with different network structures and objective functions. Besides deep models, the proposed method is also easy to be extended to protect data and traditional image processing algorithms.

An improved robust image watermarking by using different embedding strengths

Abstract: Image watermarking technique is an alternative solution to protecting digital image copyright. This paper proposed a new embedding technique based on different embedding strengths for embedding a watermark. An image is divided into non-overlapping blocks of 8 × 8 pixels. The variance pixel value was computed for each image block. Image blocks with the highest variance value were selected for the embedding regions. Therefore, it was transformed by discrete cosine transforms (DCT). Five DCT coefficients in the middle frequency were selected and the average of selected DCT blocks was calculated to generate different embedding strengths by using a set of rules. The watermark bits were embedded by using a set of embedding rules with the proposed different embedding strengths. For an additional security, the binary watermark was scrambled by using an Arnold Transform before it was embedded. The experimental results showed that the proposed scheme achieved a higher imperceptibility than the other existing schemes. The proposed scheme achieved a watermarked image quality with a PSNR value of 46 dB. The proposed scheme also produced a high watermark extracting resistance under various attacks.