AI is a transformative technology that emulates human cognitive abilities and processes large volumes of data to offer efficient solutions across various sectors of life. Although AI significantly enhances efficiency in many areas, it also presents substantial challenges, particularly regarding ethics and user privacy. These challenges are exacerbated by the inadequacy of global regulations, which may lead to potential abuse and privacy violations. This study provides an in-depth review of current AI applications, identifies future needs, and addresses emerging ethical and privacy issues. The research explores the important roles of AI technologies, including multimodal AI, natural language processing, generative AI, and deepfakes. While these technologies have the potential to revolutionize industries such as content creation and digital interactions, they also face significant privacy and ethical challenges, including the risks of deepfake abuse and the need for improved data protection through platforms like PrivAI. The study emphasizes the necessity for stricter regulations and global efforts to ensure ethical AI use and effective privacy protection. By conducting a comprehensive literature review, this research aims to provide a clear perspective on the future direction of AI and propose strategies to overcome barriers in ethical and privacy practices.

This paper explores the application of active learning to enhance machine learning classifiers for spoofing detection in automatic speaker verification (ASV) systems. Leveraging the ASVspoof 2019 database, we integrate an active learning framework with traditional machine learning workflows, specifically focusing on Random Forest (RF) and Multilayer Perceptron (MLP) classifiers. The active learning approach was implemented by initially training models on a small subset of data and iteratively selecting the most uncertain samples for further training, which allowed the classifiers to refine their predictions effectively. Experimental results demonstrate that while the MLP initially outperformed RF with an accuracy of 95.83% compared to 91%, the incorporation of active learning significantly improved RF's performance to 94%, narrowing the performance gap between the two models. After applying active learning, both classifiers showed enhanced precision, recall, and F1-scores, with improvements ranging from 3% to 5%. This study provides valuable insights into the role of active learning in boosting the efficiency of machine learning models for dynamic spoofing scenarios in ASV systems. Future research should focus on designing advanced active learning techniques and exploring their integration with other machine learning paradigms to further enhance ASV security.

Controlling Micro-Electro-Mechanical Systems (MEMS) gyroscopes often involves dealing with uncertainties and external disturbances, which can complicate control strategies. This article proposes a novel control strategy that integrates Integral Sliding Mode Control (ISMC) with Time Delay Estimation (TDE) and Arithmetic Optimization Algorithm (AOA) to enhance control performance. The proposed controller, OTDISMC, is designed to eliminate chattering and improve robustness against disturbances without relying on system dynamics. Contrary to the conventional controllers structures which depended on the system dynamic in their schemes, a model free controller is formulated without using system dynamics in its formulation. Time delay estimation technique has been undertaken as an efficient approximating strategy to approximate and compensate the lumped uncertain dynamics of the system. AOA has been undertaken to determine the optimum solutions of the coefficients of proposed control approach. The stability has been analyzed and investigated using the Lyapunov stability criterion. To show the effectiveness and validity of the developed controller, computer simulations in nominal and robustness scenarios have been carried out and compared with TDISMC that tuned by trial and error and PSO-TDISMC that tuned by particle swarm optimization (PSO). Simulation results demonstrate that OTDISMC significantly reduces tracking errors and improves robustness. The results indicate the superiority of the proposed controller as compared with traditional TDISMC tuned by classical methods and PSO-TDISMC tuned by particle swarm optimization (PSO).

The enhancement of self-driving vehicles has the potential to disrupt traditional transportation systems, Utilizing progress in secure and intelligent mobility. However, control of movement in self-driving vehicles is still difficult to carry out driving duties in a constantly changing road environment. The regulation of bearing angle is an essential component in self-driving vehicles navigation systems, facilitating the secure and efficient operation of vehicles across a range of environments, including urban streets, highways, and off-road terrain. It employs algorithms and sensor fusion to perceive surroundings, compute trajectories, and execute precise steering commands. The bearing angle represents the angle between the vehicle's current and desired directions. By consistently monitoring this angle and implementing appropriate steering inputs, the self-driving vehicle can accurately stay on track and proactively adapt to obstacles or adhere to a designated route. In this context, we explore the advancements in bearing angle control methods for self-driving vehicles. By conducting simulations of a simplified block diagram for a self-guiding vehicle's bearing angle control techniques, the efficacy of the steering system of self-driving cars has been briefly examined. We provide various methods of control, which are considered approaches for controlling the angle of bearings through lag lead compensation and PID auto-tuned controllers. The results show that the auto-tuned PID controller outperforms all other controllers in terms of transient and steady-state responses.

Rocket flight tests invariably serve a purpose, one of which involves area monitoring or aerial photography. Consequently, the rocket necessitates the installation of a camera that remains consistently oriented toward the Earth's surface throughout its trajectory. Thus, ensuring the rocket's stability and preventing any rotation becomes imperative. To achieve this, the Onboard Attitude Determination Control System (OADCS) was researched and developed, fully controlled by NI myRIO with Labview as the programming language, ensures the rocket's attitude control and maintains a rolling angle of 0 degrees during flight. The MyRIO oversees the retrieval of attitude and position data from the X-Plane flight simulator, offering feedback through actuator control. The development of the OADCS proceeded incrementally through stages utilizing the Software in the Loop Simulation (SILS) and Hardware in the Loop Simulation (HILS) techniques, to ensure the verification of the system's functionality before its application to the rocket for real flight testing. In the OADCS control scheme, Model Predictive Control (MPC) is chosen, and it is compared with a PID controller to serve as a benchmark for processing speed. Because the rocket's flight time is short and its speeds of up to Mach 4. The simulation results indicate that MPC can halt the rocket's rotation 12 times more rapidly than PID control. Additionally, the MPC's ability to maintain a zero-degree rotation can persist throughout the rocket's flight time. Employing SILS and HILS enhances the OADCS rocket development process by incorporating MPC, which holds promise for application in real rockets.

Recent developments in Wireless Sensor Networks (WSN) focus on scalability and reliability. This research addresses the challenge of improving reliability in WSNs through optimal relay placement and multipath topology design. A heuristic method with a Multi-Objective Optimization (MOO) approach is proposed to solve this problem. The proposed method integrates a modified Genetic Algorithm (GA) with Particle Swarm Optimization (PSO) characteristics. The hybrid approach aims to minimize the number of relays and associated communication costs while maintaining network reliability. The method encodes relay positions and quantities into GA chromosomes that are updated by mutation, crossover, and PSO-inspired particle motion. Simulations are performed in a simplified square area with twenty randomly placed sensors, a hundred and thirty-two arranged relays, and a single sink node. As a result, the simulation generated two multipath topologies that offer unique advantages. The first emphasizes relay efficiency (61 relays, with 2096 costs), while the second ensures lower communication costs (64 relays, 1832 costs). Comparisons with alternative algorithms, including Dijkstra, A-star, GA, and PSO, prove the superiority of the proposed approach. The optimum results obtained with a composition of 95% GA and 5% PSO, outperform alternative algorithms in terms of relay efficiency and communication cost. This research contributes to the field by providing a robust solution for designing reliable multipath WSNs with a minimum number of relays.

This research reviews AI integration in AVs, evaluating its effectiveness in urban and highway settings. Analyzing over 161 studies, it explores advancements like machine learning perception, sensor technology, V2X communication, and adaptive cruise control. It also examines challenges like traffic congestion, pedestrian and cyclist safety, regulations, and technology limitations. Safety considerations include human-AI interaction, cybersecurity, and liability/ethics. The study contributes valuable insights into the latest developments and challenges of AI in AVs, specifically in urban and highway contexts, which will guide future transportation research and decision-making. In urban settings, AI-powered sensor fusion technology helps AVs navigate dynamic traffic safely. On highways, adaptive cruise control systems maintain safe distances, reducing accidents. These findings suggest AI facilitates safer navigation in urban areas and enhances safety and efficiency on highways. While AI integration in AVs holds immense potential, innovative solutions like advanced perception systems and optimized long-range communication are needed to create safer and more sustainable transportation systems.

With the rising prevalence of retinal diseases, identifying eye diseases at an early stage is crucial for effective treatment and prevention of irreversible blindness. But Ophthalmologists face challenges in detecting subtle symptoms that may indicate the presence of a disease before it progresses to an advanced stage Among these challenges, eye diseases can present with a wide range of symptoms, and some conditions may share similar signs. To solve these difficulties, in the research proposed YOLOV8(You Only Look Once) Lightweight Self-Attention model to classify seven different retinal diseases. In this regard, the dataset that have been used in this study contains 5787 images from three different sources (Roboflow, Kaggle and Medical Clinics) were included in the seven classes of Glaucoma, Age-related Macular Degeneration (AMD), Cataract, Diabetic retinopathy (DR), and Retinal Vein Occlusion, which comprises of Branch Retinal Vein Occlusion (BRVO) and Central Retinal Occlusion (CRVO) and normal. As a results, the model has proven excellent performance in its classification ability. Boasting an average classification accuracy of 94% across the seven disease with precsition 96.2%, recall 96.6%and f1 score was 96.3% At the time of training it was 0.6 Houres(H). When compaired with Resnet50, VGG16 results underscore the model’s superior performance in precision and computational efficiency compared. The algorithm's evaluation reveals its superiority when compared to earlier pertinent research, making it a trustworthy method for classifying retinal illnesses.

This research investigates the impact of model simplification on the dynamic performance of an ABB IRB-140 six-jointed industrial robotic arm, concentrating on torque prediction and energy consumption. The entire mathematical model of forward, reverse, differential kinematics, and dynamic model proposed based on the technical specifications of the arm, and to obtain the center of the mass and inertia matrices, which are essential components of the dynamic model, Utilizing Solidworks, we developed three CAD/CAM models representing the manipulator with varying detail levels, such as simplified, semi-detailed, and detailed. Our findings indicate minor differences in the model's torque and energy consumption graphs. The semi-detailed model consumed the most energy, except for joint 1, with the detailed model showing a 0.53% reduction and the simplified model a 6.8% reduction in energy consumption. Despite these variations, all models proved effective in predicting the robot's performance during a standard 30-second task, demonstrating their adequacy for various industrial applications. This research highlights the balance between computational efficiency and accuracy in model selection. While the detailed model offers the highest precision, it demands more computational resources, which is suitable for high-precision tasks. In discrepancy, simplified, less precise models offer computational efficiency, making them adequate for specific scenarios. Our study provides critical insights into selecting dynamic models in industrial robotics. It guides the optimization of performance and energy efficiency based on the required task precision and available computational resources. This comprehensive comparison of dynamic models underscores their applicability and effectiveness in diverse industrial settings.

Aflatoxin contamination in cacao is a significant problem in terms of trade losses and health effects. This calls for the need for a non-invasive, precise, and effective detection strategy. This research contribution is to determine the best deep-learning model to classify the aflatoxin contamination level in cocoa beans based on fluorescence images and deep learning to improve performance in the classification. The process involved inoculating and incubating Aspergillus flavus (6mL/100g) to obtain aflatoxin-contaminated cocoa beans for 7 days during the incubation period. Liquid Mass Chromatography (LCMS) was used to quantify the aflatoxin in order to categorize the images into different levels including “free of aflatoxin”, “contaminated below the limit”, and “contaminated above the limit”. 300 images were acquired through a mini studio equipped with UV lamps. The aflatoxin level was classified using several pre-trained CNN approaches which has high accuracy such as GoogLeNet, SqueezeNet, AlexNet, and ResNet50. The sensitivity analysis showed that the highest classification accuracy was found in the GoogLeNet model with optimizer: Adam and learning rate: 0.0001 by 96.42%. The model was tested using a testing dataset and obtain accuracy of 96% based on the confusion matrix. The findings indicate that combining CNN with fluorescence images improved the ability to classify the amount of aflatoxin contamination in cacao beans. This method has the potential to be more accurate and economical than the current approach, which could be adapted to reduce aflatoxin's negative effects on food safety and cacao trade losses.