In this paper, we propose new 3D CNN prediction models for detecting student engagement levels in an e-learning environment. The first generated model classifies students’ engagement to high positive engagement or low positive engagement. The second generated model classifies engagement to low negative engagement or disengagement. To predict the engagement level, the proposed prediction models learn the deep spatiotemporal features of the body activities of the students. In addition, we collected a new video dataset for this study. The new dataset was collected in realistic, uncontrolled settings from real students attending real online classes. Our findings are threefold: (1) Spatiotemporal features are more suitable for analyzing body activities from video data; (2) our proposed prediction models outperform state-of-the-art methods and have proven their effectiveness; and (3) our newly collected video dataset, which reflects realistic scenarios, contributed to delivering comparable results to current methods. The findings of this work will strengthen the knowledge base for the development of intelligent and interactive e-learning systems that can give feedback based on user engagement.

Surveillance remains an important research area, and it has many applications. Smart surveillance requires a high level of accuracy even when persons are uncooperative. Gait Recognition is the study of recognizing people by the way they walk even when they are unwilling to cooperate. It is another form of a behavioral biometric system in which unique attributes of an individual’s gait are analyzed to determine their identity. On the other hand, one of the big limitations of the gait recognition system is uncooperative environments in which both gallery and probe sets are made under different and unknown walking conditions. In order to tackle this problem, we propose a deep learning‐based method that is trained on individuals with the normal walking condition, and to deal with an uncooperative environment and recognize the individual with any dynamic walking conditions, a cycle consistent generative adversarial network is used. This method translates a GEI disturbed from different covariate factors to a normal GEI. It works like unsupervised learning, and during its training, a GEI disrupts from different covariate factors of each individual and acts as a source domain while the normal walking conditions of individuals are our target domain to which translation is required. The cycle consistent GANs automatically find an individual pair with the help of the Cycle Loss function and generate the required GEI, which is tested by the CNN model to predict the person ID. The proposed system is evaluated over a publicly available data set named CASIA‐B, and it achieved excellent results. Moreover, this system can be implemented in sensitive areas, like banks, seminar halls (events), airports, embassies, shopping malls, police stations, military areas, and other public service areas for security purposes.

In recent years, various applications have emerged requiring linear topologies of wireless sensor networks (WSN). Such topologies are used in pipeline (water/oil/gas) monitoring systems. The linear structure has a significant impact on network performance in terms of delay, throughput, and power consumption. Regarding communication efficiency, routing protocols play a critical role, considering the special requirements of linear topology and energy resources. Therefore, the challenge is to design effective routing protocols that can address the diverse requirements of the monitoring system. In this paper, we present various wireless communication technologies and existing leak detection systems. We review different routing protocols focusing on multi-hop hierarchical protocols, highlighting the limitations and design issues related to packet routing in linear pipeline leak detection networks. Additionally, we present a LoRa multi-hop model for monitoring aboveground oil pipelines. A set of model parameters are identified such as the distance between sensors. In addition, the paper determines some calculations to estimate traffic congestion and energy consumption. Several alternative model designs are investigated. The model is evaluated using different multi-hop communication scenarios, and we compare the data rate and energy to provide an energy-efficient and low-cost leak detection system.

This paper focuses on the study of IoT network deployments, in both unlicensed and licensed bands, considering LoRaWAN and NB-IoT standards, respectively. The objective is to develop a comprehensive and detailed network planning and coverage dimensioning methodology for assessing key metrics related to the achieved throughput and capacity for specific requirements in order to identify tradeoffs and key issues that are related to the applicability of IoT access technologies for representative use case types. This paper will provide a concise overview of key characteristics of IoT representative IoT access network standards that are considered for being deployed in unlicensed and licensed bands and will present a methodology for modeling the characteristics of both access network technologies in order to assess their coverage and capacity considering different parameters.

Machine learning, particularly deep learning, has revolutionized a number of fields, including medical diagnostics. In this study, federated learning is employed to address privacy concerns and data access limitations inherent in medical imaging. A simulated FL environment was used to investigate the performance of five pre-trained neural network models: DENSENET121, RESNET18, VGG-NET11, GOOGLENET and INCEPTION-V3. It emphasizes the optimization of training duration as well as the application of lossy image compression techniques such as JPEG in order to improve communication efficiency. We conducted a comparative analysis of the models' performance before and after image compression by evaluating the Area Under the Receiver Operating Characteristic Curve and the training time. According to the results, image compression can maintain or improve model performance while affecting training time, underscoring the trade-offs between model accuracy and computational efficiency.