Abstract Deep learning techniques have attained substantial progress in various face‐related tasks, such as face recognition, face inpainting, and facial expression recognition. To prevent infection or the spread of the virus, wearing of masks in public places has been mandated following the COVID‐19 epidemic, which has led to face occlusion and posed significant challenges for face recognition systems. Most prominent masked face recognition solutions rely on mask segmentation tasks. Therefore, segmentation can be used to mitigate the negative impacts of wearing a mask and improve recognition accuracy. Mask region segmentation suffers from two main problems: there is no standard type of masks that people wear, they come in different colours and designs, and there is no publicly available masked face dataset with appropriate ground truth for the mask region. In order to address these issues, we propose an encoder–decoder framework that utilizes a boundary‐aware attention network combined with a new hybrid loss to provide a map, patch, and pixel‐level supervision. We also introduce a dataset called MFSD, with 11,601 images and 12,758 masked faces for masked face segmentation. Furthermore, we compare the performance of different cutting‐edge deep learning semantic segmentation models on the presented dataset. Experimental results on the MSFD dataset reveal that the suggested approach outperforms state‐of‐the‐art, algorithms with 97.623% accuracy, 93.814% IoU, and 96.817% F1‐score rate. Our dataset of masked faces with mask region labels and source code will be available online.

Abstract Segmentation is crucial in diagnosing retinal diseases by accurately identifiying retinal vessels. This paper addresses the complexity of segmenting retinal vessels, highlighting the need for precise analysis of blood vessel structures. Despite the progress made by convolutional neural networsks (CNNs) in image segmentation, their limitations in capturing the global structure of retinal vsessels and maintaining segmentation continuity present challenges. To tackle these issues, our proposed network integrates graph convolutional networks (GCNs) and attention mechansims. This allows the model to consider pixel relationships and learn vessel graphical structures, significantly improving segmentation accuracy. Additionally, the attentional feature fusion module, including pixel‐wise and channel‐wise attention mechansims within the U‐Net architecture, refines the model's focus on relevant features. This paper emphasizes the importance of continuty preservation, ensuring an accurate representation of pixel‐level information and structural details during sefmentation. Therefore, our method performs as an effective solution to overcome challenges in retinal vessel segmentation. The proposed method outperformed the state‐of‐the‐art approaches on DRIVE (Digital Retinal Images for Vessel Extraction) and STARE (Structed Analysis of the Retina) datasets with accuracies of 0.12% and 0.14%, respecttively. Importantly, our proposed approach excelled in delineating slender and diminutive blood vessels, crucial for diagnosing vascular‐related diseases. Implementation is accessible on https://github.com/CVLab‐SHUT/VGA‐Net .

Abstract A recent approach to image steganography is to use deep learning. Mainly, convolutional neural networks can extract complex features and use them as patterns to combine hidden messages and images. Also, by using generative adversarial networks, it is possible to generate realistic and high‐quality stego images without any noticeable artifacts. Previous methods suffered from challenges such as simple architecture, low network accuracy, imbalance between capacity and transparency, vanishing gradients, and low capacity. This study introduces a steganography framework named VidaGAN that utilizes deep learning techniques. The network being proposed is made up of three components: an encoder, a decoder, and a critic, and introduces a novel architecture and several innovations to address some of the unresolved challenges mentioned above. This study introduces a novel method for embedding any type of binary data into images using generative adversarial networks, enabling us to enhance the visual appeal of images generated by the specified model. This neural network called VarIable aDAptive GAN (VidaGAN) achieved state‐of‐the‐art status by reaching a hiding capacity of 3.9 bits per pixel in the DIV2K dataset. Furthermore, examination by the StegExpose steganalysis tool shows an AUC of 0.6, a suitable threshold for transparency.

Due to the increasing sophistication of web attacks, Web Application Firewalls (WAFs) have to be tested and updated regularly to resist the relentless flow of web attacks. In practice, using a brute-force attack to discover vulnerabilities is infeasible due to the wide variety of attack patterns. Thus, various black-box testing techniques have been proposed in the literature. However, these techniques suffer from low efficiency. This article presents Reinforcement-Learning-Driven and Adaptive Testing (RAT), an automated black-box testing strategy to discover injection vulnerabilities in WAFs. In particular, we focus on SQL injection and Cross-site Scripting, which have been among the top ten vulnerabilities over the past decade. More specifically,RATclusters similar attack samples together. It then utilizes a reinforcement learning technique combined with a novel adaptive search algorithm to discover almost all bypassing attack patterns efficiently. We compareRATwith three state-of-the-art me&thods considering their objectives. The experiments show thatRATperforms 33.53 and 63.16 percent on average better than its counterparts in discovering the most possible bypassing payloads and reducing the number of attempts before finding the first bypassing payload when testing well-configured WAFs, respectively.

The next generation of wireless networks including Five and Six Generations ( 5 G and 6 G ) can provide very high data rates as a demand for the Internet of Everything (IoE) system which connects millions of people and billions of machines. To reach such a high data rate, the wireless networks should work at high enough frequencies, such as millimeter and THz bands, which in turn suffer from a large attenuation and acute multipath fading. The idea of coating any objects in the environment with Software-Defined Metasurfaces (SDMs) was presented to control these effects by managing the electromagnetic properties of the environment. Since the programmable environment can be changed during the communication, for example, a sudden obstacle appears, this management should be adaptive. This paper presents the use of a reinforcement learning (RL) algorithm for dynamically configuring such an environment. In other words, when a change happens in the environment, for example, an obstacle blocks some EM waves, the agent receives a large punishment, and therefore a new action is selected. In our model, the transmitted electromagnetic waves and the tiles are considered as the agents and states, respectively. Moreover, the actions of each tile include absorbing or reflecting the impinging waves in a specific direction. We utilize the Q-learning technique to establish proper wireless links between the users and the access point (AP) by controlling the state of the tiles in an environment covered by the SDMs. Evaluation of the proposed model for different scenarios, including emerging sudden obstacles, indicates its potential to provide a proper signal level for all the users and improve the average received power up to 12 % in comparison with the related works.

LSB matching revisited is an LSB-based approach for image steganography. This method is a type of coding to increase the capacity of steganography. In this method, two bits of the secret message are hidden in two pixels with only one change. But this method provides no idea for hiding a message with a large number of bits. In other words, this method works only for n = 2 , where n is the number of bits in a block of the secret message. In this paper, we propose an improved version of the LSB matching revisited approach, which works for n > 2 . The proposed scheme contains two phases including embedding and extracting the message. In the embedding phase, we first convert the secret message into a bit-stream, and then the bit-stream is divided into a set of blocks including n bits in each block. Then we choose 2 n − 1 pixels for hiding such n bits of the secret message. In the next step, we choose the operations needed to generate such a message. Finally, we perform the obtained operations over the coefficients to hide the secret message. The proposed approach needs fewer changes than LSB MR when n > 2 . The capacity of the proposed approach is 2 n − 1 / 2 n − 1 − 1 × 100 % higher than the F5 method where this value for n > 2 is bigger than 75%. For example, the capacity of our scheme is 75% higher than the capacity of F5 for n = 3 . The proposed method can be used in the first step of every steganography method to reduce the change in the stego image. Therefore, this method is a new coding method for steganography. Our experimental results using steganalysis show that using our method provides around 10% higher detection error for SRNet over two steganography schemes.

Optical character recognition, known as OCR, has been widely used due to high demand of different technologies. Currently, most existing OCR systems have been focused on Latin languages. In recent studies, OCR systems for non‐Latin texts involving cursive style have also been introduced despite posing some challenges. In this study, the authors propose an OCR system based on long short‐term memory neural networks for the Persian language. The authors also investigate the effects of variations of parameters, involved in this approach. The proposed OCR system solves false recognition of sub‐word ‘LA’ and ‘LA’. Moreover, the authors present a preprocessing algorithm to remove ‘justification’ using image processing. A new comprehensive collated data set is introduced, comprising five million images with eight popular Persian fonts and in ten various font sizes. The proposed evaluations show that the accuracy of the proposed OCR is increased by 2%, compared to the existing Persian OCR system. The experimental results indicated that the proposed system has average accuracy of 99.69% at the letter level. The proposed system has an accuracy of 98.1% for ‘zero‐width non‐breaking space’ and 98.64% for ‘LA’ at the word level.

Brain–computer interface (BCI) is a powerful system for communicating between the brain and outside world. Traditional BCI systems work based on electroencephalogram (EEG) signals only. Recently, researchers have used a combination of EEG signals with other signals to improve the performance of BCI systems. Among these signals, the combination of EEG with functional near‐infrared spectroscopy (fNIRS) has achieved favourable results. In most studies, only EEGs or fNIRs have been considered as chain‐like sequences, and do not consider complex correlations between adjacent signals, neither in time nor channel location. In this study, a deep neural network model has been introduced to identify the exact objectives of the human brain by introducing temporal and spatial features. The proposed model incorporates the spatial relationship between EEG and fNIRS signals. This could be implemented by transforming the sequences of these chain‐like signals into hierarchical three‐rank tensors. The tests show that the proposed model has a precision of 99.6%.