Biblio

Style transfer is a research hotspot in computer vision. Up to now, it is still a challenge although many researches have been conducted on it for high quality style transfer. In this work, we propose an algorithm named ASTCNN which is a real-time Arbitrary Style Transfer Convolution Neural Network. The ASTCNN consists of two independent encoders and a decoder. The encoders respectively extract style and content features from style and content and the decoder generates the style transferred image images. Experimental results show that ASTCNN achieves higher quality output image than the state-of-the-art style transfer algorithms and the floating point computation of ASTCNN is 23.3% less than theirs.

Neural style transfer has drawn broad attention in recent years. However, most existing methods aim to explicitly model the transformation between different styles, and the learned model is thus not generalizable to new styles. We here attempt to separate the representations for styles and contents, and propose a generalized style transfer network consisting of style encoder, content encoder, mixer and decoder. The style encoder and content encoder are used to extract the style and content factors from the style reference images and content reference images, respectively. The mixer employs a bilinear model to integrate the above two factors and finally feeds it into a decoder to generate images with target style and content. To separate the style features and content features, we leverage the conditional dependence of styles and contents given an image. During training, the encoder network learns to extract styles and contents from two sets of reference images in limited size, one with shared style and the other with shared content. This learning framework allows simultaneous style transfer among multiple styles and can be deemed as a special 'multi-task' learning scenario. The encoders are expected to capture the underlying features for different styles and contents which is generalizable to new styles and contents. For validation, we applied the proposed algorithm to the Chinese Typeface transfer problem. Extensive experiment results on character generation have demonstrated the effectiveness and robustness of our method.

The following topics are dealt with: Internet of Things; invasive software; security of data; program testing; reverse engineering; product codes; binary codes; decoding; maximum likelihood decoding; field programmable gate arrays.

Intrusion detection systems (IDS) are important security tools. NIDS monitors network's traffic and HIDS filters local one. HIDS are often based on anomaly detection. Several studies deal with anomaly detection using system-call traces. In this paper, we propose an anomaly detection and prediction approach. System-call traces, invoked by the running programs, are analyzed in real time. For prediction, we use a Sequence to sequence model based on variational encoder-decoder (VED) and variants of Recurrent Neural Networks (RNN), these architectures showed their performance on natural language processing. To make the analogy, we exploit the semantics behind the invoking order of system-calls that are then seen as sentences. A preprocessing phase is added to optimize the prediction model input data representation. A one-class classification is done to categorize the sequences into normal or abnormal. Tests are achieved on the ADFA-LD dataset and showed the advantage of the prediction for the intrusion detection/prediction task.

It is notably challenging to design an efficient and secure signature scheme based on error-correcting codes. An approach to build such signature schemes is to derive it from an identification protocol through the Fiat-Shamir transform. All such protocols based on codes must be run several rounds, since each run of the protocol allows a cheating probability of either 2/3 or 1/2. The resulting signature size is proportional to the number of rounds, thus making the 1/2 cheating probability version more attractive. We present a signature scheme based on double circulant codes in the rank metric, derived from an identification protocol with cheating probability of 2/3. We reduced this probability to almost 1/2 to obtain the smallest signature among code-based signature schemes based on the Fiat-Shamir paradigm, around 22 KBytes for 128 bit security level. Furthermore, among all code-based signature schemes, our proposal has the lowest value of signature plus public key size, and the smallest secret and public key sizes. We provide a security proof in the Random Oracle Model, implementation performances, and a comparison with the parameters of similar signature schemes.

We propose a serverless computing mechanism for distributed computation based on polar codes. Serverless computing is an emerging cloud based computation model that lets users run their functions on the cloud without provisioning or managing servers. Our proposed approach is a hybrid computing framework that carries out computationally expensive tasks such as linear algebraic operations involving large-scale data using serverless computing and does the rest of the processing locally. We address the limitations and reliability issues of serverless platforms such as straggling workers using coding theory, drawing ideas from recent literature on coded computation. The proposed mechanism uses polar codes to ensure straggler-resilience in a computationally effective manner. We provide extensive evidence showing polar codes outperform other coding methods. We have designed a sequential decoder specifically for polar codes in erasure channels with full-precision input and outputs. In addition, we have extended the proposed method to the matrix multiplication case where both matrices being multiplied are coded. The proposed coded computation scheme is implemented for AWS Lambda. Experiment results are presented where the performance of the proposed coded computation technique is tested in optimization via gradient descent. Finally, we introduce the idea of partial polarization which reduces the computational burden of encoding and decoding at the expense of straggler-resilience.

Although sequence-to-sequence attentional neural machine translation (NMT) has achieved great progress recently, it is confronted with two challenges: learning optimal model parameters for long parallel sentences and well exploiting different scopes of contexts. In this paper, partially inspired by the idea of segmenting a long sentence into short clauses, each of which can be easily translated by NMT, we propose a hierarchy-to-sequence attentional NMT model to handle these two challenges. Our encoder takes the segmented clause sequence as input and explores a hierarchical neural network structure to model words, clauses, and sentences at different levels, particularly with two layers of recurrent neural networks modeling semantic compositionality at the word and clause level. Correspondingly, the decoder sequentially translates segmented clauses and simultaneously applies two types of attention models to capture contexts of interclause and intraclause for translation prediction. In this way, we can not only improve parameter learning, but also well explore different scopes of contexts for translation. Experimental results on Chinese-English and English-German translation demonstrate the superiorities of the proposed model over the conventional NMT model.

In this paper, we propose an exact closed-form expression of secrecy outage probability (SOP) for underlay cognitive network with energy scavenging capable relay over Nakagami-m fading channels and under both (maximum transmit and interference) power constraints. Various results validated the proposed expression and shed insights into the security performance of this network in key specifications.

With an increasing number of wireless devices, the risk of being eavesdropped increases as well. From information theory, it is well known that wiretap codes can asymptotically achieve vanishing decoding error probability at the legitimate receiver while also achieving vanishing leakage to eavesdroppers. However, under finite blocklength, there exists a tradeoff among different parameters of the transmission. In this work, we propose a flexible wiretap code design for Gaussian wiretap channels under finite blocklength by neural network autoencoders. We show that the proposed scheme has higher flexibility in terms of the error rate and leakage tradeoff, compared to the traditional codes.

This paper gives a concept of wireless sensor network and describe the encoding algorithm and decoding algorithm along with the implementation of LDPC code and Rateless code. Compare the performances of those two code in WSN environment by making simulation in a Rayleigh channel in matlab and derive results and conclusions from the simulation.

Polar codes are a talented coding technique with the ability to accomplish the discrete memoryless channel capacity for modern communication systems with high reliability, but it is not secured enough for such systems. A secured system counts on grouping polar codes with secret Mersenne- Twister pseudo-random number generator (MT PRNG) is presented in this paper. The proposed encoder security is deduced from the secret pre-shared initial state of MT PRNG which is considered as the crypto-system ciphering key. The generated sequences are random like and control the frozen bits' values and their indices in the polarized bit channels. When the decoding cipher key at the receiver has one-bit change from the original encoding cipher key, the receiver has an almost 0.5 BER probability. This means that the receiver, in this case, had no clue about the originally sent information data bits without prior knowledge of the utilized 232-bit ciphering key. Moreover, the security of the system can be enhanced by utilizing a pseudo-random number generator (PRBG) with longer seed to increase the system secrecy and decoding obscurity.

Ze the quality of channels into either completely noisy or noieseless channels. This paper presents extrinsic information transfer (EXIT) analysis for iterative decoding of Polar codes to reveal the mechanism of channel transformation. The purpose of understanding the transformation process are to comprehend the placement process of information bit and frozen bit and to comprehend the security standard of Polar codes. Mutual information derived based on the concept of EXIT chart for check nodes and variable nodes of low density parity check (LDPC) codes and applied to Polar codes. This paper explores the quality of the polarized channels in finite blocklength. The finite block-length is of our interest since in the fifth telecommunications generation (5G) the block length is limited. This paper reveals the EXIT curve changes of Polar codes and explores the polarization characteristics, thus, high value of mutual informations for frozen bit are needed to be detectable. If it is the other way, the error correction capability of Polar codes would be drastically decreases. These results are expected to be a reference for developments of Polar codes for 5G technologies and beyond.

Secret key generation from wireless channel is an emerging technique of physical layer security. At present, most of the secret key generation schemes use information reconciliation to obtain symmetric keys. This paper introduces a non-interactive information reconciliation scheme based on channel coding and stream encryption, and considering the error correction capability, we design a concatenated code of BCH and RS codes as channel coding. The performance of concatenated error correction code has been analyzed in this scheme. Then, we compare the concatenated code with first-level error correction code in different test environments. Extensive numerical simulations and experiments demonstrate that the decoding performance of this second-level concatenated code is better than the first-level error correction code, and it can also effectively eliminate third-party eavesdropping.

In practical wireless communication systems, the channel conditions of legitimate users can not always be better than those of eavesdroppers. This realistic fact brings the challenge for the design of secure transmission over wiretap channels which requires that the eavesdropping channel conditions should be worse than legitimate channels. In this paper, we present a robust chaos-based information masking polar coding scheme for enhancing reliability and security performances under realistic channel conditions for practical systems. In our design, we mask the original information, wherein the masking matrix is determined by chaotic sequences. Then the masked information is encoded by the secure polar coding scheme. After the channel polarization achieved by the polar coding, we could identify the bit-channels providing good transmission conditions for legitimate users and the bit-channels with bad conditions for eavesdroppers. Simulations are performed over the additive white Gaussian noise (AWGN) and slow flat-fading Rayleigh channels. The results demonstrate that compared with existing schemes, the proposed scheme can achieve better reliability and security even when the eavesdroppers have better channel conditions than legitimate users, hence the practicability is greatly enhanced.

Accountable authority identity-based encryption (A-IBE), as an attractive way to guarantee the user privacy security, enables a malicious private key generator (PKG) to be traced if it generates and re-distributes a user private key. Particularly, an A-IBE scheme achieves full black-box security if it can further trace a decoder box and is secure against a malicious PKG who can access the user decryption results. In PKC'11, Sahai and Seyalioglu presented a generic construction for full black-box A-IBE from a primitive called dummy identity-based encryption, which is a hybrid between IBE and attribute-based encryption (ABE). However, as the complexity of ABE, their construction is inefficient and the size of private keys and ciphertexts in their instantiation is linear in the length of user identity. In this paper, we present a new efficient generic construction for full black-box A-IBE from a new primitive called token-based identity-based encryption (TB-IBE), without using ABE. We first formalize the definition and security model for TB-IBE. Subsequently, we show that a TB-IBE scheme satisfying some properties can be converted to a full black-box A-IBE scheme, which is as efficient as the underlying TB-IBE scheme in terms of computational complexity and parameter sizes. Finally, we give an instantiation with the computational complexity as O(1) and the constant size master key pair, private keys, and ciphertexts.

We introduce a novel distributed derivative-free optimization framework that is resilient to stragglers. The proposed method employs coded search directions at which the objective function is evaluated, and a decoding step to find the next iterate. Our framework can be seen as an extension of evolution strategies and structured exploration methods where structured search directions were utilized. As an application, we consider black-box adversarial attacks on deep convolutional neural networks. Our numerical experiments demonstrate a significant improvement in the computation times.

The application of deep recurrent networks to audio transcription has led to impressive gains in automatic speech recognition (ASR) systems. Many have demonstrated that small adversarial perturbations can fool deep neural networks into incorrectly predicting a specified target with high confidence. Current work on fooling ASR systems have focused on white-box attacks, in which the model architecture and parameters are known. In this paper, we adopt a black-box approach to adversarial generation, combining the approaches of both genetic algorithms and gradient estimation to solve the task. We achieve a 89.25% targeted attack similarity, with 35% targeted attack success rate, after 3000 generations while maintaining 94.6% audio file similarity.

Consider the problem of guessing a random vector X by submitting queries (guesses) of the form "Is X equal to x?" until an affirmative answer is obtained. A key figure of merit is the number of queries required until the right vector is guessed, termed the guesswork. The goal is to devise a guessing strategy which minimizes a certain guesswork moment. We study a universal, decentralized scenario where the guesser does not know the distribution of X, and is not allowed to prepare a list of words to be guessed in advance, or to remember its past guesses. Such a scenario is useful, for example, if bots within a Botnet carry out a brute-force attack to guess a password or decrypt a message, yet cannot coordinate the guesses or even know how many bots actually participate in the attack. We devise universal decentralized guessing strategies, first, for memoryless sources, and then generalize them to finite-state sources. For both, we derive the guessing exponent and prove its asymptotic optimality by deriving a matching converse. The strategies are based on randomized guessing using a universal distribution. We also extend the results to guessing with side information (SI). Finally, we design simple algorithms for sampling from the universal distributions.

In the era of digitization, data security is a vital role in message transmission and all systems that deal with users require stronger encryption techniques that against brute force attack. Honey encryption (HE) algorithm is a user data protection algorithm that can deceive the attackers from unauthorized access to user, database and websites. The main part of conventional HE is distribution transforming encoder (DTE). However, the current DTE process using cumulative distribution function (CDF) has the weakness in message space limitation because CDF cannot solve the probability theory in more than four messages. So, we propose a new method in DTE process using discrete distribution function in order to solve message space limitation problem. In our proposed honeywords generation method, the current weakness of existing honeywords generation method such as storage overhead problem can be solved. In this paper, we also describe the case studies calculation of DTE in order to prove that new DTE process has no message space limitation and mathematical model using discrete distribution function for DTE process facilitates the distribution probability theory.

In today's world, securing data is becoming one of the main issues, the elaboration of the fusion of cryptography and steganography are contemplating as the sphere of on-going research. This can be gain by cryptography, steganography, and fusion of these two, where message firstly encoding using any cryptography techniques and then conceal into any cover medium using steganography techniques. Biological structure of DNA is used as the cover medium due to high storage capacity, simple encoding method, massive parallelism and randomness DNA cryptography can be used in identification card and tickets. Currently work in this field is still in the developmental stage and a lot of investigation is required to reach a fully-fledged stage. This paper provides a review of the existing method of DNA based cryptography

We consider a one-time digital signature scheme recently proposed by Persichetti and show that a successful key recovery attack can be mounted with limited complexity. The attack we propose exploits a single signature intercepted by the attacker, and relies on a statistical analysis performed over such a signature, followed by information set decoding. We assess the attack complexity and show that a full recovery of the secret key can be performed with a work factor that is far below the claimed security level. The efficiency of the attack is motivated by the sparsity of the signature, which leads to a significant information leakage about the secret key.

The paper deals with the design and principles of functioning of code-based schemes for formation and verification of electronic digital signature. Comparative studies of the effectiveness of the known CFS scheme and the proposed scheme have been carried out, as well as their possibilities, disadvantages and prospects for use in the post-quantum period.

Cyber-physical systems (CPS) are state-of-the-art communication environments that offer various applications with distinct requirements. However, security in CPS is a nonnegotiable concept, since without a proper security mechanism the applications of CPS may risk human lives, the privacy of individuals, and system operations. In this paper, we focus on PHY-layer security approaches in CPS to prevent passive eavesdropping attacks, and we propose an integration of physical layer operations to enhance security. Thanks to the McEliece cryptosystem, error injection is firstly applied to information bits, which are encoded with the forward error correction (FEC) schemes. Golay and Hamming codes are selected as FEC schemes to satisfy power and computational efficiency. Then obtained codewords are transmitted across reliable intermediate relays to the legitimate receiver. As a performance metric, the decoding frame error rate of the eavesdropper is analytically obtained for the fragmentary existence of significant noise between relays and Eve. The simulation results validate the analytical calculations, and the obtained results show that the number of low-quality channels and the selected FEC scheme affects the performance of the proposed model.

Multiple-input multiple-output (MIMO) techniques are currently the de facto approach for increasing the capacity and reliability of communication systems. Spatial modulation (SM) is presently one of the most eminent MIMO techniques. As, it combines the advantages of having higher spectral efficiency than repetition coding (RC) while overcoming the inter-channel interference (ICI) faced by spatial multiplexing (SMP). Moreover, SM reduces system complexity. In this paper, for the first time in literature, the use of MIMO techniques is explored in Internet-of-Things(IoT) deployments by introducing a novel technique called security aware spatial modulation (SA-SM).SA-SM provides a low complexity, secure and spectrally efficient technique that harvests the advantages of SM, while facing the arising security concerns of IoT systems. Using an undemanding modification at the receiver, SA-SM gives an extra degree of technology independent physical layer security. Our results show that SA-SM forces the bit-error-rate (BER) of an eavesdropper to not exceed the range of 10-2, which is below the forward-error-correction (FEC) threshold. Hence, it eradicates the ability of an eavesdropper to properly decode the transmitted signal. Additionally, the efficiency of SA-SM is verified in both the radio and visible light ranges. Furthermore, SA-SM is capable of reducing the peak-to-average-power-ratio (PAPR) by 26.2%.