Biblio

The fourth industrial revolution has led to the rapid development of industrial control systems. While the large number of industrial system devices connected to the Internet provides convenience for production management, it also exposes industrial control systems to more attack surfaces. Under the influence of multiple attack surfaces, sensitive data leakage has a more serious and time-spanning negative impact on industrial production systems. How to quickly locate the source of information leakage plays a crucial role in reducing the loss from the attack, so there are new requirements for tracing sensitive data in industrial control information systems. In this paper, we propose a digital watermarking traceability scheme for sensitive data in industrial control systems to address the above problems. In this scheme, we enhance the granularity of traceability by classifying sensitive data types of industrial control systems into text, image and video data with differentiated processing, and achieve accurate positioning of data sources by combining technologies such as national secret asymmetric encryption and hash message authentication codes, and mitigate the impact of mainstream watermarking technologies such as obfuscation attacks and copy attacks on sensitive data. It also mitigates the attacks against the watermarking traceability such as obfuscation attacks and copy attacks. At the same time, this scheme designs a data flow watermark monitoring module on the post-node of the data source to monitor the unauthorized sensitive data access behavior caused by other attacks.

In this paper, we propose a novel watermarking-based copy deterrence scheme for identifying data leaks through authorized query users in secure image outsourcing systems. The scheme generates watermarks unique to each query user, which are embedded in the retrieved encrypted images. During unauthorized distribution, the watermark embedded in the image is extracted to determine the untrustworthy query user. Experimental results show that the proposed scheme achieves minimal information loss, faster embedding and better resistance to JPEG compression attacks compared with the state-of-the-art schemes.

With the rapid development of multimedia and short video, there is a growing concern for video copyright protection. Some work has been proposed to add some copyright or fingerprint information to the video to trace the source of the video when it is stolen and protect video copyright. This paper proposes a video watermarking method based on a deep neural network and curriculum learning for watermarking of sliced videos. The first frame of the segmented video is perturbed by an encoder network, which is invisible and can be distinguished by the decoder network. Our model is trained and tested on an online educational video dataset consisting of 2000 different video clips. Experimental results show that our method can successfully discriminate most watermarked and non-watermarked videos with low visual disturbance, which can be achieved even under a relatively high video compression rate(H.264 video compress with CRF 32).

Machine learning (ML) models are increasingly being used in the development of Malware Detection Systems. Existing research in this area primarily focuses on developing new architectures and feature representation techniques to improve the accuracy of the model. However, recent studies have shown that existing state-of-the art techniques are vulnerable to adversarial machine learning (AML) attacks. Among those, data poisoning attacks have been identified as a top concern for ML practitioners. A recent study on clean-label poisoning attacks in which an adversary intentionally crafts training samples in order for the model to learn a backdoor watermark was shown to degrade the performance of state-of-the-art classifiers. Defenses against such poisoning attacks have been largely under-explored. We investigate a recently proposed clean-label poisoning attack and leverage an ensemble-based Nested Training technique to remove most of the poisoned samples from a poisoned training dataset. Our technique leverages the relatively large sensitivity of poisoned samples to feature noise that disproportionately affects the accuracy of a backdoored model. In particular, we show that for two state-of-the art architectures trained on the EMBER dataset affected by the clean-label attack, the Nested Training approach improves the accuracy of backdoor malware samples from 3.42% to 93.2%. We also show that samples produced by the clean-label attack often successfully evade malware classification even when the classifier is not poisoned during training. However, even in such scenarios, our Nested Training technique can mitigate the effect of such clean-label-based evasion attacks by recovering the model's accuracy of malware detection from 3.57% to 93.2%.

A dual-image watermarking approach is presented in this research. The presented work utilizes the properties of Hessenberg decomposition, Redundant discrete wavelet transform (RDWT), Discrete cosine transform (DCT) and Singular value decomposition (SVD). For watermarking, the YCbCr color space is employed. Two watermark logos are for embedding. A YCbCr format conversion is performed on the RGB input image. The host image's Y and Cb components are divided into various sub-bands using RDWT. The Hessenberg decomposition is applied on high-low and low-high components. After that, SVD is applied to get dominant matrices. Two different logos are used for watermarking. Apply RDWT on both watermark images. After that, apply DCT and SVD to get dominant matrices of logos. Add dominant matrices of input host and watermark images to get the watermarked image. Average PSNR, MSE, Structural similarity index measurement (SSIM) and Normalized correlation coefficient (NCC) are used as the performance parameters. The resilience of the presented work is tested against various attacks such as Gaussian low pass filter, Speckle noise attack, Salt and Pepper, Gaussian noise, Rotation, Median and Average filter, Sharpening, Histogram equalization and JPEG compression. The presented scheme is robust and imperceptible when compared with other schemes.

Watermarking is one of the most common data hiding techniques for multimedia elements. Broadcasting, copy control, copyright protection and authentication are the most frequently used application areas of the watermarking. Secret data can be embedded into the cover image with changing the values of the pixels in spatial domain watermarking. In addition to this method, cover image can be converted into one of the transformation such as Discrete Wavelet Transformation (DWT), Discrete Cousin Transformation (DCT) and Discrete Fourier Transformation (DFT). Later on watermark can be embedded high frequencies of transformation coefficients. In this work, cover image transformed one, two and three level DWT decompositions. Binary watermark is hided into the low and high frequencies in each decomposition. Experimental results show that watermarked image is robust, secure and resist against several geometric attacks especially JPEG compression, Gaussian noise and histogram equalization. Peak Signal-to-Noise Ratio (PSNR) and Similarity Ratio (SR) values show very optimal results when we compare the other frequency and spatial domain algorithms.

Image watermarking techniques provides security, reliability copyright protection for various multimedia contents. In this paper Integer Wavelet Transform Schur decomposition and Singular value decomposition (SVD) based image watermarking scheme is suggested for the integrity protection of dicom images. In the proposed technique 3-level Integer wavelet transform (IWT) is subjected into the Dicom ultrasound image of liver cover image and in HH sub-band Schur decomposition is applied. The upper triangular matrix obtained from Schur decomposition of HH sub-band is further processed with SVD to attain the singular values. The X-ray watermark image is pre-processed before embedding into cover image by applying 3-level IWT is applied into it and singular matrix of LL sub-band is embedded. The watermarked image is encrypted using Arnold chaotic encryption for its integrity protection. The performance of suggested scheme is tested under various attacks like filtering (median, average, Gaussian) checkmark (histogram equalization, rotation, horizontal and vertical flipping, contrast enhancement, gamma correction) and noise (Gaussian, speckle, Salt & Pepper Noise). The proposed technique provides strong robustness against various attacks and chaotic encryption provides integrity to watermarked image.

In the present era of the internet, image watermarking schemes are used to provide content authentication, security and reliability of various multimedia contents. In this paper image watermarking scheme which utilizes the properties of Integer Wavelet Transform (IWT), Schur decomposition and Singular value decomposition (SVD) based is proposed. In the suggested method, the cover image is subjected to a 3-level Integer wavelet transform (IWT), and the HH3 subband is subjected to Schur decomposition. In order to retrieve its singular values, the upper triangular matrix from the HH3 subband’s Schur decomposition is then subjected to SVD. The watermark image is first encrypted using a chaotic map, followed by the application of a 3-level IWT to the encrypted watermark and the usage of singular values of the LL-subband to embed by manipulating the singular values of the processed cover image. The proposed scheme is tested under various attacks like filtering (median, average, Gaussian) checkmark (histogram equalization, rotation, horizontal and vertical flipping) and noise (Gaussian, Salt & Pepper Noise). The suggested scheme provides strong robustness against numerous attacks and chaotic encryption provides security to watermark.

Nowadays, data is essential in several fields, such as science, finance, medicine, and transportation, which means its value continues to rise. Relational databases are vulnerable to copyright threats when transmitted and shared as a carrier of data. The watermarking technique is seen as a partial solution to the problem of securing copyright ownership. However, most of them are currently restricted to numerical attributes in relational databases, limiting their versatility. Furthermore, they modify the source data to a large extent, failing to keep the characteristics of the original database, and they are susceptible to solid malicious attacks. This paper proposes a new robust reversible watermarking technique, Fields Based Inserting Position Tuples algorithm (FBIPT), for relational databases. FBIPT does not modify the original database directly; instead, it inserts some position tuples based on three Fields―Group Field, Feature Field, and Control Field. Field information can be calculated by numeric attributes and any attribute that can be transformed into binary bits. FBIPT technique retains all the characteristics of the source database, and experimental results prove the effectiveness of FBIPT and show its highly robust performance compared to state-of-the-art watermarking schemes.

Automatic Identification System (AIS) plays a leading role in maritime navigation, traffic control, local and global maritime situational awareness. Today, the reliable and secure AIS operation is threatened by probable cyber attacks such as imitation of ghost vessels, false distress or security messages, or fake virtual aids-to-navigation. We propose a method for ensuring the authentication and integrity of AIS messages based on the use of the Message Authentication Code scheme and digital watermarking (WM) technology to organize an additional tag transmission channel. The method provides full compatibility with the existing AIS functionality.

The wide application of deep learning techniques is boosting the regulation of deep learning models, especially deep neural networks (DNN), as commercial products. A necessary prerequisite for such regulations is identifying the owner of deep neural networks, which is usually done through the watermark. Current DNN watermarking schemes, particularly white-box ones, are uniformly fragile against a family of functionality equivalence attacks, especially the neuron permutation. This operation can effortlessly invalidate the ownership proof and escape copyright regulations. To enhance the robustness of white-box DNN watermarking schemes, this paper presents a procedure that aligns neurons into the same order as when the watermark is embedded, so the watermark can be correctly recognized. This neuron alignment process significantly facilitates the functionality of established deep neural network watermarking schemes.

Image steganography is the operation of hiding a message into a cover image. the message can be text, codes, or image. Hiding an image into another is the proposed approach in this paper. Based on LSB coding, a k-LSB-based method is proposed using k least bits to hide the image. For decoding the hidden image, a region detection operation is used to know the blocks contains the hidden image. The resolution of stego image can be affected, for that, an image quality enhancement method is used to enhance the image resolution. To demonstrate the effectiveness of the proposed approach, we compare it with some of the state-of-the-art methods.

In recent years, coverless image steganography attracts significant attentions due to its distortion-free trait on carrier images to avoid the detection by steganalysis tools. Despite this advantage, current coverless methods face several challenges, e.g., vulnerability to geometrical attacks and low hidden capacity. In this paper, we propose a novel coverless steganography algorithm based on chaotic encrypted dual radial harmonic Fourier moments (DRHFM) to tackle the challenges. In specific, we build mappings between the extracted DRHFM features and secret messages. These features are robust to various of attacks, especially to geometrical attacks. We further deploy the DRHFM parameters to adjust the feature length, thus ensuring the high hidden capacity. Moreover, we introduce a chaos encryption algorithm to enhance the security of the mapping features. The experimental results demonstrate that our proposed scheme outperforms the state-of-the-art coverless steganography based on image mapping in terms of robustness and hidden capacity.

Many techniques are available nowadays, for Intellectual Property(IP) protection of Digital circuits. Out of these techniques, the popular one is watermarking. Similar to the watermarking used in case of text, image and video, watermarking of digital circuits also modifies a digital circuit design in such a way, that only the IP owner of design is able to extract the watermark form the design. In this paper, Multi – Feedback configuration of Linear Feedback Shift Register(LFSR) is used to watermark a FSM based design. This watermarking technique improves the watermark strength of already existing LFSR based watermarking technique. In terms of hardware utilization, it is significantly efficient than some popular watermarking techniques. The proposed technique has been implemented using Verilog HDL in Xilinx ISE and the simulation is done using ModelSim.

In the last decade the use of artificial neural networks (ANNs) in many fields like image processing or speech recognition has become a common practice because of their effectiveness to solve complex tasks. However, in such a rush, very little attention has been paid to security aspects. In this work we explore the possibility to embed a watermark into the ANN parameters. We exploit model redundancy and adaptation capacity to lock a subset of its parameters to carry the watermark sequence. The watermark can be extracted in a simple way to claim copyright on models but can be very easily attacked with model fine-tuning. To tackle this culprit we devise a novel watermark aware training strategy. We aim at delving into the loss landscape to find an optimal configuration of the parameters such that we are robust to fine-tuning attacks towards the watermarked parameters. Our experimental results on classical ANN models trained on well-known MNIST and CIFAR-10 datasets show that the proposed approach makes the embedded watermark robust to fine-tuning and compression attacks.

Sharing relational databases on the Internet creates the need to protect these databases. Its output in substantial losses to the data storing systems because of unauthorized access to information that could lose novelty. The research associations use the research databases to mine new information about the research works of the relational databases that are available for free. It is a great challenge to maintain authenticity because these databases are vulnerable to security issues. Watermarking is a candidate solution that fully protects databases shared with the receiver. The protection of relational database ownership that may continue to evolve against the various aquatic mechanisms shared with the recipient that arouses appetite for attacks and must continue to evolve so that they can have database knowledge to support their decision-making system is effective. The relational database based onVirtual private key Watermarking using numeric attribute) involves embedding the same watermark in the same properties in different places in the same place. Therefore, data attackers cannot remove watermarks from data. The proposed strategy is to work by inserting watermark bits in such a way that it causes minimal distortion in the data and the data usability must remain intact after the data is watermarked. The proposed strategy is to work by inserting watermark bits in such a way that it causes minimal distortion in the data and the ability to use the data after watermarking the data must remain intact. The existence of a primary key is the main feature or compulsory item for most of the strategies. Our method provides solutions no primary key feature where the integrating search system of the database remains intact after watermarking distortion.

In this paper, we propose a zero-distortion blind reversible robust scheme for watermarking relational databases called PKMark. Data owner can declare the copyright of the databases or pursue the infringement by extracting the water-mark information embedded in the database. PKMark is mainly based on the primary key attribute of the tuple. So it does not depend on the type of the attribute, and can provide high-precision numerical attributes. PKMark uses RSA encryption on the watermark before embedding the watermark to ensure the security of the watermark information. Then we use RSA to sign the watermark cipher text so that the owner can verify the ownership of the watermark without disclosing the watermark. The watermark embedding and extraction are based on the hash value of the primary key, so the scheme has blindness and reversibility. In other words, the user can obtain the watermark information or restore the original database without comparing it to the original database. Our scheme also has almost excellent robustness against addition attacks, deletion attacks and alteration attacks. In addition, PKMark is resistant to additive attacks, allowing different users to embed multiple watermarks without interfering with each other, and it can indicate the sequence of watermark embedding so as to indicate the original copyright owner of the database. This watermarking scheme also allows data owners to detect whether the data has been tampered with.

This research presented a digital watermarking scheme using multi-level authentication for protecting QR code images in order to provide security and authenticity. This research focuses on the improved digital watermarking scheme for QR code security that can protect the confidentiality of the information stored in QR code images from the public. Information modification, malicious attack, and copyright violation may occur due to weak security and disclosure pattern of QR code. Digital watermarking can be a solution to reduce QR code imitation and increase QR code security and authenticity. The objectives of this research are to provide QR code image authentication and security, tamper localization, and recovery scheme on QR code images. This research proposed digital watermarking for QR code images based on multi-level authentication with Least Significant Bit (LSB) and SHA-256 hash function. The embedding and extracting watermark utilized region of Interest (ROI) and Region of Non-Interest (RONI) in the spatial domain for improving the depth and width of QR code application in the anti-counterfeiting field. The experiments tested the reversibility and robustness of the proposed scheme after a tempered watermarked QR code image. The experimental results show that the proposed scheme provides multi-level security, withstands tampered attacks and it provided high imperceptibility of QR code image.

This paper considers methods of fingerprint protection in biometric authentication systems. Including methods of protecting fingerprint templates using zero digital watermarks and cryptography techniques. The paper considers a secure authentication model using cryptography and digital watermarks.

Aggressive time-to-market constraints and enormous hardware design and fabrication costs have pushed the semiconductor industry toward hardware Intellectual Properties (IP) core design. However, the globalization of the integrated circuits (IC) supply chain exposes IP providers to theft and illegal redistribution of IPs. Watermarking and fingerprinting are proposed to detect IP piracy. Nevertheless, they come with additional hardware overhead and cannot guarantee IP security as advanced attacks are reported to remove the watermark, forge, or bypass it. In this work, we propose a novel methodology, GNN4IP, to assess similarities between circuits and detect IP piracy. We model the hardware design as a graph and construct a graph neural network model to learn its behavior using the comprehensive dataset of register transfer level codes and gate-level netlists that we have gathered. GNN4IP detects IP piracy with 96% accuracy in our dataset and recognizes the original IP in its obfuscated version with 100% accuracy.

Ever since Machine Learning as a Service emerges as a viable business that utilizes deep learning models to generate lucrative revenue, Intellectual Property Right (IPR) has become a major concern because these deep learning models can easily be replicated, shared, and re-distributed by any unauthorized third parties. To the best of our knowledge, one of the prominent deep learning models - Generative Adversarial Networks (GANs) which has been widely used to create photorealistic image are totally unprotected despite the existence of pioneering IPR protection methodology for Convolutional Neural Networks (CNNs). This paper therefore presents a complete protection framework in both black-box and white-box settings to enforce IPR protection on GANs. Empirically, we show that the proposed method does not compromise the original GANs performance (i.e. image generation, image super-resolution, style transfer), and at the same time, it is able to withstand both removal and ambiguity attacks against embedded watermarks. Codes are available at https://github.com/dingsheng-ong/ipr-gan.

This paper reviews the video watermarking schemes resistance against tampering attacks. There are several transform methods which are used for Video Watermarking including Discrete Fourier Transform (DFT), Discrete Cosine Transform (DCT), Discrete wavelet transform (DWT) and are discussed and compared in this paper. The results are presented in a table with a summary of their advantages.

Digital watermarking is the multimedia leading security protection as it permanently escorts the digital content. Image copyright protection is becoming more anxious as the new 5G technology emerged. Protecting images with a robust scheme without distorting them is the main trade-off in digital watermarking. In this paper, a watermarking scheme based on discrete cosine transform (DCT) and singular value decomposition (SVD) using canny edge detector technique is proposed. A binary encrypted watermark is reshaped into a vector and inserted into the edge detected vector from the diagonal matrix of the SVD of DCT DC and low-frequency coefficients. Watermark insertion is performed by using an edge-tracing mechanism. The scheme is evaluated using the Peak Signal to Noise Ratio (PSNR) and Normalized Correlation (NC). Attained results are competitive when compared to present works in the field. Results show that the PSNR values vary from 51 dB to 55 dB.

Federated learning is a distributed learning technique where machine learning models are trained on client devices in which the local training data resides. The training is coordinated via a central server which is, typically, controlled by the intended owner of the resulting model. By avoiding the need to transport the training data to the central server, federated learning improves privacy and efficiency. But it raises the risk of model theft by clients because the resulting model is available on every client device. Even if the application software used for local training may attempt to prevent direct access to the model, a malicious client may bypass any such restrictions by reverse engineering the application software. Watermarking is a well-known deterrence method against model theft by providing the means for model owners to demonstrate ownership of their models. Several recent deep neural network (DNN) watermarking techniques use backdooring: training the models with additional mislabeled data. Backdooring requires full access to the training data and control of the training process. This is feasible when a single party trains the model in a centralized manner, but not in a federated learning setting where the training process and training data are distributed among several client devices. In this paper, we present WAFFLE, the first approach to watermark DNN models trained using federated learning. It introduces a retraining step at the server after each aggregation of local models into the global model. We show that WAFFLE efficiently embeds a resilient watermark into models incurring only negligible degradation in test accuracy (-0.17%), and does not require access to training data. We also introduce a novel technique to generate the backdoor used as a watermark. It outperforms prior techniques, imposing no communication, and low computational (+3.2%) overhead$^\textrm1$$^\textrm1$\$The research report version of this paper is also available in https://arxiv.org/abs/2008.07298, and the code for reproducing our work can be found at https://github.com/ssg-research/WAFFLE.

Recent advances in natural language generation have introduced powerful language models with high-quality output text. However, this raises concerns about the potential misuse of such models for malicious purposes. In this paper, we study natural language watermarking as a defense to help better mark and trace the provenance of text. We introduce the Adversarial Watermarking Transformer (AWT) with a jointly trained encoder-decoder and adversarial training that, given an input text and a binary message, generates an output text that is unobtrusively encoded with the given message. We further study different training and inference strategies to achieve minimal changes to the semantics and correctness of the input text.AWT is the first end-to-end model to hide data in text by automatically learning -without ground truth- word substitutions along with their locations in order to encode the message. We empirically show that our model is effective in largely preserving text utility and decoding the watermark while hiding its presence against adversaries. Additionally, we demonstrate that our method is robust against a range of attacks.