No abstract

<p><span lang="EN-US">The fast development in deep learning techniques, besides the wide spread of social networks, facilitated fabricating and distributing images and videos without prior knowledge. This paper developed an evolutionary learning algorithm to automatically design a convolutional neural network (CNN) architecture for deepfake detection. Genetic algorithm (GA) based on residual network (ResNet) and densely connected convolutional network (DenseNet) as building block units for feature extraction versus multilayer perceptron (MLP), random forest (RF) and support vector machine (SVM) as classifiers generates different CNN structures. A local search mutation operation proposed to optimize three layers: (batch normlization, activation function, and regularizes). This method has the advantage of working on different datasets without preprocessing. Findings using two datasets evidence the efficiency of the suggested approach where the generated models outperform the state-of-art by increasing 1% in the accuracy; this confirms that intuitive design is the new direction for better generalization.</span></p>

Image classification is the process of assigning labeling to the input images to a fixed set of categories; however, assigning labels to the image is difficult by using the traditional method because of the large number of images. To solve this problem, we will resort to deep learning techniques. Which is enables computers to recognize and extract visual characteristics. The convolutional neural network (CNN) is a deep neural network used for many purposes, such as image classification, detection, and face recognition, due to its high-performance accuracy in classification and detection tasks. In this paper, we develop CNN based on the transfer learning approach for image classification. The network comprises two types of transfer learning, ResNet and DenseNet, as building blocks of the network with an multilayer perceptron (MLP) classifier. The proposed method does not need to preprocess before these datasets that input into the network. It was train on two datasets: the Cifar-10 and the Sign-Traffic datasets. We conclude that the proposed method achieves the best performance compared with other states of the art. The accuracy gained is 97.45% and 99.45%, respectively, where the proposed CNN increased the accuracy compared to other methods by 3%.

Wireless video surveillance systems (WVSS) are deployed in large environments for use in strategic places such as town centers, public streets, and airports and play an essential role in protecting critical infrastructure. However, WVSSs are vulnerable to unauthorized access due to weak login credentials, which leads to their exploitation to launch cyberattacks on other systems, such as distributed denial-of-service attacks. Hence, it is essential to secure these systems from unauthorized access. This paper proposes the Mamdani fuzzy inference system (FIS)-based password checker algorithm to estimate the password strength ratio (PSR) of internet protocol (IP) cameras and internet of things (IoT) devices. This algorithm composes three stages, the password extraction stage, which evaluates the input parameters of FIS from the real-time streaming protocol (RTSP) protocol using a counter of password characters. Then, the processing stage uses Mamdani FIS to optimize the input parameters to calculate the PSR. Finally, the alarm stage will notify the system administrator about weak IoT nodes. Unlike the existing approaches, this algorithm improves detection accuracy by informing the system administrator about threatened nodes. Extensive experiments are carried out to determine the efficiency of the proposed algorithm. The results confirm the efficiency of the proposed algorithm with high accuracy, which outperforms existing schemes.

With the development of network and communication systems in large areas in the world, this leads to increase security problems in transmission of data such as data leakage, modification, unauthorized access, and attacks. There are many types of techniques that are used to prevent these problems and protect data. One of these techniques is a stream cipher which considered the strongest and fastest method used in encryption and decryption process. In this study presented a new design for the stream cipher to protect mobile data. The strength of stream cipher depends on it is' key. There are several methods to generate key. We used three types of generator. Then, it used the combiner to convert them into a nonlinear Boolean function in order to make the generator key more secure. To implement a new generator key by using these three kinds, we used four LFSRs and one of NLFSRs or FCSRs to produce five variables Boolean function. These variables will be as an input to the combiner function. Finally, we tested the generator and submitted it to the randomness tests that is publicly available in the National Institute of Standards and Technology (NIST).