A hand gesture recognition system provides a robust and innovative solution to nonverbal communication through human–computer interaction. Deep learning models have excellent potential for usage in recognition applications. To overcome related issues, most previous studies have proposed new model architectures or have fine-tuned pre-trained models. Furthermore, these studies relied on one standard dataset for both training and testing. Thus, the accuracy of these studies is reasonable. Unlike these works, the current study investigates two deep learning models with intermediate layers to recognize static hand gesture images. Both models were tested on different datasets, adjusted to suit the dataset, and then trained under different methods. First, the models were initialized with random weights and trained from scratch. Afterward, the pre-trained models were examined as feature extractors. Finally, the pre-trained models were fine-tuned with intermediate layers. Fine-tuning was conducted on three levels: the fifth, fourth, and third blocks, respectively. The models were evaluated through recognition experiments using hand gesture images in the Arabic sign language acquired under different conditions. This study also provides a new hand gesture image dataset used in these experiments, plus two other datasets. The experimental results indicated that the proposed models can be used with intermediate layers to recognize hand gesture images. Furthermore, the analysis of the results showed that fine-tuning the fifth and fourth blocks of these two models achieved the best accuracy results. In particular, the testing accuracies on the three datasets were 96.51%, 72.65%, and 55.62% when fine-tuning the fourth block and 96.50%, 67.03%, and 61.09% when fine-tuning the fifth block for the first model. The testing accuracy for the second model showed approximately similar results.

Computer security depends mainly on passwords to protect human users from attackers. Therefore, manual and alphanumerical passwords are the most frequent type of computer authentication. However, creating these passwords has significant drawbacks. For example, users often tend to choose passwords based on personal information so that they can be memorable and therefore weak and guessable. In contrast, it is often difficult to remember if the password is difficult to guess. We propose an intelligent security model for password generation and estimation to address these problems using the ensemble learning approach and hand gesture features. This paper proposes two intelligent stages: the first is the password generation stage based on the ensemble learning approach and the proposed S-Box. The second is the password strength estimation stage, also based on the ensemble learning approach. Four well-known classifiers are used: Multi-Layer Perceptron (MLP), Support Vector Machine (SVM), Random Forest Tree (RFT), and AdaBoost applied on two datasets: MNIST images dataset and password strength dataset. The experimental results showed that the hand gesture and password strength classification processes accurately performed at 99% in AUC, Accuracy, F1-measures, Precision, and Recall. As a result, the extracted features of hand gestures will directly impact the complexity of generated passwords, which are very strong, hard to guess, and memorable.

Providing ownership on relational database is a crucial issue in today internet-based application environments and in many content distribution applications.This paper provides the effective watermarking technique to protect valuable numeric relational data from illegal duplications and redistributions as well as to claim ownership,the robustness of proposed system depending on using new hybrid techniques ,first technique MAC(Message Authentication Code) that used one way hash function SHA1 ,second technique is threshold generator base on simple combination of odd number of register and by using secret key in proposed system. Detecting the watermark neither requires access to neither the original data nor the watermark. The watermark can be detected even in a small subset of a watermarked relation as long as the sample contains some of the marks. The finally stage is the analysis of technique that used, our extensive analysis shows that the proposed technique is robust against various forms of malicious attacks and updates to the data

Biometrics is a field that uses behavioral and biological traits to identify/verify a person. Characteristics include handwrittien signature, iris, gait, and fingerprint. Signature-based biometric systems are common due to their simple collection and non-intrusive. Identify the humans using their handwritten signatures has received an important attention in several modern crucial applications such as in automatic bank check, law-enforcements, and historical documents processing. Therefore, in this paper an accurate handwritten signatures system is proposed. The system uses a proposed preprocessing stage for the input handwritten signatures images. Besides, a new deep learning model called MobileNets, which used for classification process. Support vector machine (SVM) used as a classifier with the MobileNets inorder to get a better identifaction results. Experimental results conducted on standard CEDAR, ICDER, sigcomp handwritten signature datasets report 99.8%, 98.2%, 99.5%, identification accuracy, respectively.

Ease of access to digital images and the many images editing programs available, like photoshop. All this makes the Issue of protecting images against modification becomes essential. Some images contain crucial information that can risk a patient's life, such as medical images and e-government images that relate to citizen information and state or ministry security. The watermark was one of the essential methods for this type of protection, especially the fragile watermark, which is very sensitive to any attack. Because of its other characteristics, it was one of the techniques that proved its efficiency in detecting tampering and the authenticity of images—also, watermarking focuses on protecting the image itself, not about protecting the secret message. A fragile watermark is a watermarking which inserts some information to cover an image to secure it .fragile watermarking could use in such a way and implement in spatial or frequency domain or in both so, making it a hybrid watermarking scheme. The Paper presented set of fragile watermark techniques used by the researchers with the performance metrics of an algorithm used in spatial and frequency domains, also showing how to use artificial intelligence with a watermarking technique to protect Document images from manipulation and forgery. Index Terms— fragile watermark, spatial domain, frequncy domain

Recent developments in technology and the area of digital communications have rendered digital images increasingly susceptible to tampering and alteration by persons who are not authorized to do so. This may appear to be acceptable, especially if an image editing process is necessary to delete or add a particular scene that improves the quality the image. But what about images used in authorized governmental transactions? The consequences can be severe; any altered document is considered forged under the law and may cause confusion. Also, any document that cannot be verified as being authentic is regarded as a fake and cannot be used, inflicting harm on people. The suggested work intends to reduce fraud in electronic documents using a novel generator. It employs modern encryption methods to conceal the data represented by the watermark and hide it secretly inside the document so that it is easy to check for signs of counterfeiting; thus, allowing for the discovery of forged papers