Protecting an organization from numerous threats both inside and outside is the primary function of the security system. Automated embedded systems have come a long way in the contemporary era and have shown to be highly beneficial in applications related to security and surveillance. Face recognition is one of the study fields in computer vision, which is commonly used in security systems for video surveillance. Even though facial recognition technology has advanced significantly and is employed in several significant applications, numerous challenges need to be solved. These challenges include changes in posture, occlusions, expression, aging, lighting, and other elements. Deep learning can be useful in these situations. By using several processing layers to develop data representations with multiple feature extraction layers, deep learning may achieve higher accuracy. With the purpose of providing better security for student residence halls, in this work, we present a real-time deep learning-based facial recognition system that can be used to identify an individual's identity and give a warning when the individual's face is not recognized in front of the system. Here, faces from the face database are matched in order to identify students based on a video of their arrival into the residence halls. This process begins with face detection and ends with face recognition. We used a Convolutional Neural Network (CNN) based model Multi-Task Cascaded Convolutional Neural Networks (MTCNN) for face detection and recognizing faces using the Google FaceNet model. The model was trained on around 3000 photos, taken by 30 distinct people.

Protecting personal information from unauthorized access is a critical concern for individuals. However, the accumulation of confidential information by various organizations, such as banks and hospitals, for regular communication creates a potential vulnerability. If an individual visits two hospitals and both facilities independently release the individual's gathered data, a malicious adversary could potentially deduce confidential information through a composition attack. Therefore, developing methods that protect individuals from composition attacks is crucial. According to the size of the dataset and the percentage of overlapping persons, our study examines the effectiveness of composition attacks. We propose a knowledge domain-based design to mitigate successful composition attacks, which has shown promising results in reducing such attacks and compared to existing studies based on the k-anonymity and l-diversity models. Our approach leverages a knowledge domain to reduce the likelihood of data breaches, demonstrating the effectiveness of our method in protecting individuals' privacy and preventing unauthorized access to sensitive information. Finally, the effects of data utility on the diverse data set have been measured.

One of the most prevalent complications of diabetes and a leading cause of preventable blindness in the world is DR. This paper examines the history of the DR detection methods, the transition of traditional image-processing algorithms to the state-of-the-art deep learning and ensemble models. This study employed MESSIDOR dataset; it includes retinal fundus images with labels indicating DR severity. Images were enhanced and pre-processed, and extracting features based on deep learning was done using pre-trained convolutional neural networks (CNNs). In particular, the models that were used to extract the features included ResNet50, InceptionV3, DenseNet121, DenseNet169, VGG16, Xception, and a custom Deep Neural Network (DNN). Moreover, ensemble methods were designed through integrating CNN-based feature extractors with machine learning classifiers such as Random Forest, Support Vector Machine (SVM), XGBoost and LightGBM to improve the accuracy of classification and generalization. The results of the experiments showed that among the autonomous deep learning models, Xception achieved the highest performance, with a testing accuracy of 95%. EfficientNetB3 combined with XGBoost achieved the highest testing accuracy (86.83%). This study highlights the possibilities of deploying deep learning and ensemble AIs into DR screening devices, especially in telemedicine systems and mobile diagnostic systems. Such integration could significantly improve early detection rates, reduce the burden on healthcare providers, and make diabetic eye care more accessible in underserved regions.