Biblio

Similar to other digital assets, deep neural network (DNN) models could suffer from piracy threat initiated by insider and/or outsider adversaries due to their inherent commercial value. DNN watermarking is a promising technique to mitigate this threat to intellectual property. This work focuses on black-box DNN watermarking, with which an owner can only verify his ownership by issuing special trigger queries to a remote suspicious model. However, informed attackers, who are aware of the watermark and somehow obtain the triggers, could forge fake triggers to claim their ownerships since the poor robustness of triggers and the lack of correlation between the model and the owner identity. This consideration calls for new watermarking methods that can achieve better trade-off for addressing the discrepancy. In this paper, we exploit frequency domain image watermarking to generate triggers and build our DNN watermarking algorithm accordingly. Since watermarking in the frequency domain is high concealment and robust to signal processing operation, the proposed algorithm is superior to existing schemes in resisting fraudulent claim attack. Besides, extensive experimental results on 3 datasets and 8 neural networks demonstrate that the proposed DNN watermarking algorithm achieves similar performance on functionality metrics and better performance on security metrics when compared with existing algorithms.

The security of deep learning becomes increasing important with the more and more related applications. The adversarial attack is the known method that makes the performance of deep learning network (DNN) decline rapidly. However, adversarial attack needs the gradient knowledge of the target networks to craft the specific adversarial examples, which is the white-box attack and hardly becomes true in reality. In this paper, we implement a black-box attack on DNN classifier via a functionally equivalent network without knowing the internal structure and parameters of the target networks. And we increase the entropy of the noise via deep convolution generative adversarial networks (DCGAN) to make it seems fuzzier, avoiding being probed and eliminated easily by adversarial training. Experiments show that this method can produce a large number of adversarial examples quickly in batch and the target network cannot improve its accuracy via adversarial training simply.

At present, the on-site safety problems of substations and critical power equipment are mainly through inspection methods. Still, manual inspection is difficult, time-consuming, and uninterrupted inspection is not possible. The current safety management is mainly guaranteed by rules and regulations and standardized operating procedures. In the on-site environment, it is very dependent on manual execution and confirmation, and the requirements for safety supervision and operating personnel are relatively high. However, the reliability, the continuity of control and patrol cannot be fully guaranteed, and it is easy to cause security vulnerabilities and cause security accidents due to personnel slackness. In response to this shortcoming, this paper uses edge computing and image processing techniques to discover security risks in time and designs a deep convolution attention mechanism network to perform image processing. Then the network is cropped and compressed so that it can be processed at the edge, and the results are aggregated to the cloud for unified management. A comprehensive security assessment module is designed in the cloud to conduct an overall risk assessment of the results reported by all edges, and give an alarm prompt. The experimental results in the real environment show the effectiveness of this method.

This paper presents a new algorithm for image encryption by extending the Knight's Tour Problem (KTP). The idea behind the proposed algorithm is to generate a Knight Tour (KT) matrix (m,n) and then divide the image according to the size of knight tour matrix into several sub matrices. Finally, apply n-neighborhood addition modulo encryption algorithm according to the solution of KT matrix over each m × n partition of the image. The proposed algorithm provides image encryption without using the cover images. Results obtained from experiments have shown that the proposed algorithm is efficient, simple and does not disclose any information from encrypted image.

A THz image edge detection approach based on wavelet and neural network is proposed in this paper. First, the source image is decomposed by wavelet, the edges in the low-frequency sub-image are detected using neural network method and the edges in the high-frequency sub-images are detected using wavelet transform method on the coarsest level of the wavelet decomposition, the two edge images are fused according to some fusion rules to obtain the edge image of this level, it then is projected to the next level. Afterwards the final edge image of L-1 level is got according to some fusion rule. This process is repeated until reaching the 0 level thus to get the final integrated and clear edge image. The experimental results show that our approach based on fusion technique is superior to Canny operator method and wavelet transform method alone.

Since the last few decades, several chaotic encryption techniques are reported by different researchers. Although the cryptanalysis of some techniques shows the feebler resistance of those algorithms against any weaker attackers. However, different hyper-chaotic based and DNA-coding based encrypting methods are introduced recently. Though, these methods are efficient against several attacks, but, increase complexity as well. On account of these drawbacks, we have proposed a novel technique of chaotic encryption of an image using the integer wavelet transform (IWT) and global bit scrambling (GBS). Here, the image is transformed and decomposed by IWT. Thereafter, a chaotic map is used in the encryption algorithm. A key-dependent bit scrambling (GBS) is introduced rather than pixel scrambling to make the encryption stronger. It enhances key dependency along with the increased resistance against intruder attacks. To check the fragility and dependability of the algorithm, a sufficient number of tests are done, which have given reassuring results. Some tests are done to check the similarity between the original and decrypted image to ensure the excellent outcome of the decryption algorithm. The outcomes of the proposed algorithm are compared with some recent works' outputs to demonstrate its eligibility.

Hiding information is a process to hide data or include it in different digital media such as image, audio, video, and text. However, there are many techniques to achieve the process of hiding information in the image processing, in this paper, a new method has been proposed for hidden data mechanism (which is a text file), then a transposition cipher method has been employed for encryption completed. It can be used to build an encrypted text and also to increase security against possible attacks while sending it over the World Wide Web. A genetic algorithm has been affected in the adjustment of the encoded text and DNA in the creation of an encrypted text that is difficult to detect and then include in the image and that affected the image visual quality. The proposed method outperforms the state of arts in terms of efficiently retrieving the embedded messages. Performance evaluation has been recorded high visual quality scores for the (SNR (single to noise ratio), PSNR (peak single to noise ratio) and MSE (mean square error).

For real-time applications of arbitrary style transformation, there is a trade-off between the quality of results and the running time of existing algorithms. Hence, it is required to maintain the equilibrium of the quality of generated artwork with the speed of execution. It's complicated for the present arbitrary style-transformation procedures to preserve the structure of content-image while blending with the design and pattern of style-image. This paper presents the implementation of a network using SANET models for generating impressive artworks. It is flexible in the fusion of new style characteristics while sustaining the semantic-structure of the content-image. The identity-loss function helps to minimize the overall loss and conserves the spatial-arrangement of content. The results demonstrate that this method is practically efficient, and therefore it can be employed for real-time fusion and transformation using arbitrary styles.

One of the significant advancements in information technology is Cloud computing, but the security issue of data storage is a big problem in the cloud environment. That is why a system is proposed in this paper for improving the security of cloud data using encryption, information concealment, and hashing functions. In the data encryption phase, we implemented hybrid encryption using the algorithm of AES symmetric encryption and the algorithm of RSA asymmetric encryption. Next, the encrypted data will be hidden in an image using LSB algorithm. In the data validation phase, we use the SHA hashing algorithm. Also, in our suggestion, we compress the data using the LZW algorithm before hiding it in the image. Thus, it allows hiding as much data as possible. By using information concealment technology and mixed encryption, we can achieve strong data security. In this paper, PSNR and SSIM values were calculated in addition to the graph to evaluate the image masking performance before and after applying the compression process. The results showed that PSNR values of stego-image are better for compressed data compared to data before compression.

Image Encryption is a technique where an algorithm along with a set of characters called key encrypts the data into cipher text. The cipher text can be converted back into plaintext by decryption. This technique is employed for the security of data such that confidentiality, integrity and authenticity of data is maintained. In today's era security of information has become a crucial task, unauthorized access and use of data has become a noticeable issue. To provide the security required, there are several algorithms to suit the purposes. While the use and transferring of images has become easy and faster due to technological advancements especially wireless sensor network, image destruction and illegitimate use has become a potential threat. Different transfer mediums and various uses of images require different and appropriately suiting encryption approaches. Hence, in this paper we discuss the types of image encryption techniques. We have also discussed several encryption algorithms, their advantages and suitability.

An acoustic fingerprint is a condensed and powerful digital signature of an audio signal which is used for audio sample identification. A fingerprint is the pattern of a voice or audio sample. A large number of algorithms have been developed for generating such acoustic fingerprints. These algorithms facilitate systems that perform song searching, song identification, and song duplication detection. In this study, a comprehensive and powerful survey of already developed algorithms is conducted. Four major music fingerprinting algorithms are evaluated for identifying and analyzing the potential hurdles that can affect their results. Since the background and environmental noise reduces the efficiency of music fingerprinting algorithms, behavioral analysis of fingerprinting algorithms is performed using audio samples of different languages and under different environmental conditions. The results of music fingerprint classification are more successful when deep learning techniques for classification are used. The testing of the acoustic feature modeling and music fingerprinting algorithms is performed using the standard dataset of iKala, MusicBrainz and MIR-1K.

New attacks on existing algorithms have revealed various weaknesses. To overcome these weaknesses, researchers are searching for new ones. In this study, a new image encryption algorithm has been proposed to realize this aim. The most obvious component of the proposed image encryption algorithm is the cryptographic substitution box structure used to provide the confusion requirement. The original aspect of the substitution box structure according to similar studies in the literature is that the design was realized by using Lotka-Volterra chaotic system. The analysis results showed that both the substitution box structure and the image encryption algorithm provide the necessary criteria for data security.

Chaos-based encryption is one of the promising cryptography techniques that can be used. Although chaos-based encryption provides excellent security, the finite precision of number representation in computers affects decryption accuracy negatively. In this paper, a way to mitigate some problems regarding finite precision is analyzed. We show that the use of maps with small Lyapunov exponents can improve the performance of chaotic encryption scheme, making it suitable for image encryption.

Dynamic random growth technique and a hybrid chaotic map which is proposed in this paper are used to perform block-based image encryption. The plaintext attack can easily crack the cat map, as it is periodic, and therefore cat map securely used in which it can eliminate the cyclical occurrence and withstand the plaintext attack's effect. The diffusion process calculates the intermediate parameters according to the image block. For the generation of the random data stream in the chaotic map, we use an intermediate parameter as an initial parameter. In this way, the generated data stream depends on the plain text image that can withstand the attack on plain text. The experimental results of this process prove that the proposed dynamic random growth technique and a hybrid chaotic map for image encryption is a secured one in which it can be used in secured image transmission systems.

Human action recognition in video is one of the most widely applied topics in the field of image and video processing, with many applications in surveillance (security, sports, etc.), activity detection, video-content-based monitoring, man-machine interaction, and health/disability care. Action recognition is a complex process that faces several challenges such as occlusion, camera movement, viewpoint move, background clutter, and brightness variation. In this study, we propose a novel human action recognition method using convolutional neural networks (CNN) and deep bidirectional LSTM (DB-LSTM) networks, using only raw video frames. First, deep features are extracted from video frames using a pre-trained CNN architecture called ResNet152. The sequential information of the frames is then learned using the DB-LSTM network, where multiple layers are stacked together in both forward and backward passes of DB-LSTM, to increase depth. The evaluation results of the proposed method using PyTorch, compared to the state-of-the-art methods, show a considerable increase in the efficiency of action recognition on the UCF 101 dataset, reaching 95% recognition accuracy. The choice of the CNN architecture, proper tuning of input parameters, and techniques such as data augmentation contribute to the accuracy boost in this study.

Style transfer is an emerging trend in the fields of deep learning's applications, especially in images and audio data this is proven very useful and sometimes the results are astonishing. Gradually styles of textual data are also being changed in many novel works. This paper focuses on the transfer of the sentimental vibe of a sentence. Given a positive clause, the negative version of that clause or sentence is generated keeping the context same. The opposite is also done with negative sentences. Previously this was a very tough job because the go-to techniques for such tasks such as Recurrent Neural Networks (RNNs) [1] and Long Short-Term Memories(LSTMs) [2] can't perform well with it. But since newer technologies like Generative Adversarial Network(GAN) and Variational AutoEncoder(VAE) are emerging, this work seem to become more and more possible and effective. In this paper, Multi-Genarative Variational Auto-Encoder is employed to transfer sentiment values. Inspite of working with a small dataset, this model proves to be promising.

Most of the notable artworks of all time are hand drawn by great artists. But, now with the advancement in image processing and huge computation power, very sophisticated synthesised artworks are being produced. Since mid-1990's, computer graphics engineers have come up with algorithms to produce digital paintings, but the results were not visually appealing. Recently, neural networks have been used to do this task and the results seen are like never before. One such algorithm for this purpose is the neural style transfer algorithm, which imparts the pattern from one image to another, producing marvellous pieces of art. This research paper focuses on the roles of various parameters involved in the neural style transfer algorithm. An extensive analysis of how these parameters influence the output, in terms of time, performance and quality of the style transferred image produced is also shown in the paper. A concrete comparison has been drawn on the basis of different time and performance metrics. Finally, optimal values for these discussed parameters have been suggested.

Mobile expressive rendering gained increasing popularity among users seeking casual creativity by image stylization and supports the development of mobile artists as a new user group. In particular, neural style transfer has advanced as a core technology to emulate characteristics of manifold artistic styles. However, when it comes to creative expression, the technology still faces inherent limitations in providing low-level controls for localized image stylization. This work enhances state-of-the-art neural style transfer techniques by a generalized user interface with interactive tools to facilitate a creative and localized editing process. Thereby, we first propose a problem characterization representing trade-offs between visual quality, run-time performance, and user control. We then present MaeSTrO, a mobile app for orchestration of neural style transfer techniques using iterative, multi-style generative and adaptive neural networks that can be locally controlled by on-screen painting metaphors. At this, first user tests indicate different levels of satisfaction for the implemented techniques and interaction design.

The notion of integrated circuit split manufacturing which delegates the front-end-of-line (FEOL) and back-end-of-line (BEOL) parts to different foundries, is to prevent overproduction, piracy of the intellectual property (IP), or targeted insertion of hardware Trojans by adversaries in the FEOL facility. In this work, we challenge the security promise of split manufacturing by formulating various layout-level placement and routing hints as vector- and image-based features. We construct a sophisticated deep neural network which can infer the missing BEOL connections with high accuracy. Compared with the publicly available network-flow attack [1], for the same set of ISCAS-85benchmarks, we achieve 1.21× accuracy when splitting on M1 and 1.12× accuracy when splitting on M3 with less than 1% running time.

Most of the security algorithms proposed for the sensor networks such as secure routing, data encryption and authentication, and intrusion detection target protecting the content of the collected data from being exposed to different types of attacks. However, the context of the collected data, such as event occurrence, event time, and event location, is not addressed by these security mechanisms and can still be leaked to the adversaries. Therefore, we propose in this paper a novel and efficient unobservability scheme for source/sink location privacy for wireless multimedia sensor networks. The proposed privacy scheme is based on a cross-layer design between the application and routing layers in order to exploit the multimedia processing technique with multipath routing to hide the event occurrences and locations of important nodes without degrading the network performance. Simulation analysis shows that our proposed scheme satisfies the privacy requirements and has better performance compared to other existing techniques.

In this paper, a novel method for tomographic microwave imaging based on the Compressive Processing (CP) paradigm is proposed. The retrieval of the dielectric profiles of the scatterers is carried out by efficiently solving both the sampling and the sensing problems suitably formulated under the first order Born approximation. Selected numerical results are presented in order to show the improvements provided by the CP with respect to conventional compressive sensing (CSE) approaches.

Biometric recognition systems based on ultrasound images have been investigated for several decades, and nowadays ultrasonic fingerprint sensors are fully integrated in portable devices. Main advantage of the Ultrasound over other technologies are the possibility to collect 3D images, allowing to gain information on under-skin features, which improve recognition accuracy and resistance to spoofing. Also, ultrasound images are not sensible to several skin contaminations, humidity and not uniform ambient illumination. An ultrasound system, able to acquire 3D images of the human palm has been recently proposed. In this work, a recognition procedure based on 2D palmprint images collected with this system is proposed and evaluated through verification experiments carried out on a home made database composed of 141 samples collected from 24 users. Perspective of the proposed method by upgrading the recognition procedure to provide a 3D template able to accounts for palm lines' depth are finally highlighted and discussed.

Recently in the vast advancement of Artificial Intelligence, Machine learning and Deep Neural Network (DNN) driven us to the robust applications. Such as Image processing, speech recognition, and natural language processing, DNN Algorithms has succeeded in many drawbacks; especially the trained DNN models have made easy to the researchers to produces state-of-art results. However, sharing these trained models are always a challenging task, i.e. security, and protection. We performed extensive experiments to present some analysis of watermark in DNN. We proposed a DNN model for Digital watermarking which investigate the intellectual property of Deep Neural Network, Embedding watermarks, and owner verification. This model can generate the watermarks to deal with possible attacks (fine tuning and train to embed). This approach is tested on the standard dataset. Hence this model is robust to above counter-watermark attacks. Our model accurately and instantly verifies the ownership of all the remotely expanded deep learning models without affecting the model accuracy for standard information data.

As the accessibility of high-resolution smartphone camera has increased and an improved computational speed, it is now convenient to build Business Card Readers on mobile phones. The project aims to design and develop a Business Card Reader (BCR) Application for iOS devices, using an open-source OCR Engine - Tesseract. The system accuracy was tested and evaluated using a dataset of 55 digital business cards obtained from an online repository. The accuracy result of the system was up to 74% in terms of both text recognition and data detection. A comparative analysis was carried out against a commercial business card reader application and our application performed vastly reasonable.

Digital image security is now a severe issue, especially when sending images to telecommunications networks. There are many ways where digital images can be encrypted and decrypted from secure communication. Digital images contain data that is important when captured or disseminated to preserve and preserve data. The technique of encryption is one way of providing data on digital images. A key cipher block chaining and Gingerbreadman Map are used in our search to encrypt images. This new system uses simplicity, high quality, enhanced by the vehicle's natural efficiency and the number of the chain. The proposed method is performed for experimental purposes and the experiments are performed in- depth, highly reliable analysis. The results confirm that by referring to several known attacks, the plan cannot be completed. Comparative studies with other algorithms show a slight rise in the security of passwords with the advantages of security of the chain. The results of this experiment are a comparison of button sensitivity and a comparison after encryption and decryption of the initial image using the amount of pixel change rate and unified average change intensity.