Human Action Recognition (HAR) in unconstrained video remains difficult due to cluttered backgrounds, camera motion, and long-range temporal dependencies. The recognition of human action is the most complex study in the area of Artificial Intelligence (AI) and Computer Vision (CV). Machine vision for online and offline video processing is typically employed in the development of human behavior recognition systems. In video broadcasting and analysis, identifying the type and content of human actions present in the footage is a fundamental requirement. In this article, we propose a compact and deployable Convolutional Neural Network-Long Short Term Memory (CNN-LSTM) framework that combines a 2... More >

The proliferation of unmanned aerial vehicles (UAVs) has driven the emergence of Flying Ad-hoc Networks (FANETs) as distributed cyber-physical computing systems for surveillance, disaster management, precision agriculture, and logistics. Optimizing computational and communication resource utilization within these constrained aerial systems is essential to extend network lifespan and maintain satisfactory quality of service (QoS). This study proposes integrating the Fisheye State Routing (FSR) protocol with a 3$\times$3 Manhattan Grid mobility model as a computing-systems solution for reliable data exchange in dynamic UAV swarm networks. FSR maintains distributed topology tables and applies... More >

Vietnam's Electric Vehicle (EV) market is expanding rapidly, yet public charging infrastructure development lags significantly, exhibiting pronounced spatial imbalance in dense urban cores. This study addresses this gap through an integrated demand forecasting and location optimization framework for District 1, Ho Chi Minh City. We develop a log-linear regression model using Vietnam's macroeconomic data (2003–2023), identifying GDP and CPI as dominant determinants of vehicle ownership (R$^2$ = 0.962). Forecasted vehicle stocks for 2026–2030 are translated into public charging demand through vehicle-type disaggregation and service-capacity modeling. Spatially, we propose a four-stage opti... More >

AI-driven metaheuristic optimization algorithms have demonstrated strong potential in solving complex, high-dimensional engineering design problems. This paper applies and systematically compares three representative swarm-intelligence and evolutionary computation methods—particle swarm optimization (PSO), whale optimization algorithm (WOA), and grey wolf optimizer (GWO)—to the parameter design of LCL filters in single-phase grid-tied inverter systems, a class of problems characterized by nonlinear constraints and competing performance objectives. A structured optimization framework is developed, incorporating an objective function targeting total harmonic distortion (THD) minimization a... More >

In today's technological landscape, recommender systems provide essential personalized suggestions by leveraging user preferences. This study evaluates the computational performance of User-Based (UBCF) and Model-Based Collaborative Filtering (MBCF) as intelligent computing systems on the MovieLens 1M dataset, comparing performance on complete data versus partitions based on age and occupation. Using MAE and RMSE metrics with an 80/20 train-test split, we assessed UBCF with Euclidean/Cosine similarity and MBCF with NMF/SVD. Results show MBCF with SVD achieved the best performance (MAE: 0.6909, RMSE: 0.8761), outperforming UBCF by approximately 5.2% in MAE and 5.1% in RMSE (p $<$ 0.05). Thi... More >

The growth of the Internet of Things (IoT) has connected a massive number of devices, but its common centralized design creates major security, privacy, and scalability problems that old security methods cannot properly fix. This review explores how Blockchain Technology (BCT) offers a new approach to create trust and strong security for IoT systems without a central authority. Following the PRISMA guidelines, this study analyzes 86 research papers and uses a Chi-square test to statistically confirm the link between IoT problems and the use of blockchain solutions. The results show a strong, statistically proven connection, with a Chi-square value of 34.772 (p<0.05) and a Cramer's V of 0.26.... More >

Gaze estimation plays a vital role in human-computer interaction, driver monitoring, and psychological analysis. While state-of-the-art appearance-based methods achieve high accuracy using deep learning, they often demand substantial computational resources, including GPU acceleration and extensive training, limiting their use in resource-constrained or real-time scenarios. This paper introduces GeoGaze, a novel, lightweight, training-free framework that infers categorical gaze direction (“Left”, “Center”, “Right”) solely from geometric analysis of facial landmarks. Leveraging the high-precision 478-point face mesh and iris landmarks provided by MediaPipe, GeoGaze computes a simp... More >

Text-based CAPTCHAs remain a widely deployed mechanism to distinguish humans from automated bots. The Telerik RadCaptcha, a component of the ASP.NET AJAX suite, generates distorted alphanumeric images with character overlap, intersecting lines, and dynamic background noise. This study introduces a novel, real-world dataset of 3,000 labeled Telerik RadCaptcha images and proposes a specialized multi-stage preprocessing pipeline featuring adaptive binarization and contour-based segmentation to robustly isolate overlapping and noisy characters—challenges where conventional methods frequently fail. The segmented characters are then classified using a lightweight Convolutional Neural Network (CN... More >