IJSRET Volume 10 Issue 6, Nov-Dec-2024

Abstract-Generative Artificial Intelligence (GenAI) has evolved rapidly, creating transformative opportunities across sectors, particularly in healthcare and marketing. Despite the promise of improved patient care, streamlined medical workflows, and enhanced customer engagement, GenAI faces significant challenges. Key obstacles include high computational costs, data-privacy concerns, and ethical accountability in content generation. Moreover, the open-source initiatives by leading firms like Meta have intensified competition, pushing GenAI models toward commoditization, impacting revenue structures and sparking a “race to the bottom” in pricing. The market is further complicated by monopolistic dependencies on critical hardware providers, particularly Nvidia, which dominate GPU supplies essential for AI training. With a rapidly growing market projected to reach trillions by 2030, the industry must navigate these barriers to realize the full potential of GenAI. This study explores GenAI’s current applications, fiscal and ethical challenges, and the strategic imperatives needed to foster sustainable, profitable growth within an increasingly crowded and commoditized industry landscape.

Abstract-In the era of digital transformation, organizations are increasingly leveraging data analytics to optimize marketing strategies and enhance customer engagement. Predicting customer performance is critical for businesses aiming to tailor marketing efforts, improve customer retention, and maximize revenue. This study presents a comprehensive data mining framework utilizing the Naive Bayes classifier to forecast customer performance based on historical behavior and interaction data. Employing Google Analytics as the primary data collection tool, we evaluate the model’s effectiveness by analyzing metrics such as accuracy, True Positive Rate (TPR), False Positive Rate (FPR), and the area under the Receiver Operating Characteristic (ROC) curve. The results illustrate the framework’s potential to provide actionable insights into customer behavior, thereby facilitating more informed marketing strategies and decision-making processes.

Abstract-IIn the current times, conventional farming which is the most widely used type of farming has been affected by several problems such as decrease in the availability of space due to the increasing population, wastage of water, destruction of crops due to insects, rains, etc. Furthermore, in the future where the population is expected to grow further, these problems in farming can be disastrous as it can decrease the availability of food and can lead to the starvation of a big part of the population. Hydroponics which is another method of farming can be a solution to most of the problems associated with conventional farming. In this type of farming, crops are grown without the requirement of soil, instead it utilizes a growing medium and water is directly supplied to the roots of the plants. Further fertilizers are dissolved in the water itself. This type of farming can save a lot of space as the plants are grown in vertical slots and they can be stacked upon each other and water requirement is also very low for this type of farming as most of the water is recycled. In this paper, we are going to discuss the various factors which affect the growth rate of the plants in vertical farming. The plants we have taken are jalapeno plants. The trail period is of 7 weeks where we have compared different factors affecting the growth rate of the plants.

Abstract-Chatbots function as software that allows users to ask questions and receive assistance through appropriate responses. This paper explores an AI-based chatbot designed specifically for students experiencing suicidal thoughts or at risk of suicide. The aim of this chatbot is to help reduce the number of suicides among students by providing them with timely support and guidance. Leveraging the expansive and rapidly evolving field of AI, this technology can contribute positively to addressing societal challenges and promoting well-being.

Abstract-As data demands increase with the evolution toward beyond-5G and 6G communication systems, achieving efficient network backhaul is crucial to support high data rates, minimized latency, and broad geographic coverage. Traditional backhaul networks, reliant on radio frequency (RF) communications, face limitations in scalability and bandwidth, particularly in dense urban and rural remote areas. This paper explores a hybrid RF-Free-Space Optical (FSO) communication model, integrating Low Earth Orbit (LEO) satellites with High Altitude Platform Stations (HAPS) to enhance backhaul network efficiency. The proposed HAPS-LEO cooperative model mitigates atmospheric disruptions and offers scalable, high-bandwidth solutions. We further examine Contact Graph Routing (CGR) as a protocol for optimized data routing in variable connectivity conditions, presenting simulated performance results that demonstrate the advantages of this architecture.

Abstract-Cardiovascular diseases (CVDs) are a leading cause of death worldwide, posing a significant health threat not only in India but across the globe. This highlights the critical need for a dependable, precise, and accessible system to diagnose such conditions promptly, enabling timely treatment. Machine learning algorithms have become invaluable tools in healthcare, automating the analysis of extensive and complex datasets. Recent studies demonstrate that various machine learning techniques can aid healthcare professionals in diagnosing heart-related conditions. The heart, second only to the brain in importance, plays a vital role in circulating blood throughout the body. Predicting heart disease occurrence is thus essential in the medical field. Data analytics enhances the prediction accuracy by analysing large volumes of patient data, often maintained on a monthly basis, which could be utilized to anticipate potential future diseases. Techniques such as Artificial Neural Networks (ANN), Random Forest, and Support Vector Machines (SVM) are widely applied to predict heart conditions. Diagnosing and predicting heart diseases remain a considerable challenge for both doctors and hospitals globally. To mitigate the high mortality rate associated with these diseases, efficient and rapid detection methods are essential. Machine learning and data mining techniques hold a crucial role in this context. Researchers are accelerating efforts to develop machine learning-based software that can assist doctors in both predicting and diagnosing heart diseases. This research project aims to leverage machine learning algorithms to predict the likelihood of heart disease in patients.

Abstract-This project focuses on traffic safety analysis, aiming to enhance road user safety through a comprehensive evaluation of various factors that influence accident rates and driving conditions. By assessing parameters such as skid resistance, surface texture, visibility, and roadway geometry, the study identifies critical factors that contribute to traffic incidents and offers insights into effective safety measures. Field data was gathered from selected road sections, and laboratory tests were conducted to analyze surface characteristics. Statistical analysis was applied to understand the correlation between these factors and accident frequency, enabling the development of targeted recommendations to improve safety standards. The project underscores the importance of proactive road maintenance and design improvements in reducing accidents and enhancing the overall safety and efficiency of transportation infrastructure. This project aims to enhance road safety by conducting an in-depth analysis of factors contributing to traffic accidents and assessing the effectiveness of potential interventions. Through examining elements such as pavement skid resistance, surface texture, road geometry, and visibility, the study explores their influence on accident frequency and severity. Field data collected from high-risk road sections, along with laboratory testing of pavement properties, provide a foundation for evaluating existing conditions. Using statistical and spatial analysis, the study identifies patterns in accident data, highlighting critical areas for improvement. Recommendations are developed based on these insights to propose cost-effective strategies that prioritize safety, such as optimized pavement materials, better signage, and improved road design. This research underscores the role of systematic traffic safety analysis in advancing safer, more resilient transportation systems. This project undertakes a comprehensive traffic safety analysis aimed at reducing accidents and improving road safety through a detailed examination of key factors affecting driving conditions. By focusing on parameters such as skid resistance, pavement surface texture, visibility, road geometry, and traffic flow, the study seeks to identify elements that significantly impact accident rates and driving safety.

Abstract-This project focuses on traffic safety analysis, aiming to enhance road user safety through a comprehensive evaluation of various factors that influence accident rates and driving conditions. By assessing parameters such as skid resistance, surface texture, visibility, and roadway geometry, the study identifies critical factors that contribute to traffic incidents and offers insights into effective safety measures. Field data was gathered from selected road sections, and laboratory tests were conducted to analyze surface characteristics. Statistical analysis was applied to understand the correlation between these factors and accident frequency, enabling the development of targeted recommendations to improve safety standards. The project underscores the importance of proactive road maintenance and design improvements in reducing accidents and enhancing the overall safety and efficiency of transportation infrastructure. This project aims to enhance road safety by conducting an in-depth analysis of factors contributing to traffic accidents and assessing the effectiveness of potential interventions. Through examining elements such as pavement skid resistance, surface texture, road geometry, and visibility, the study explores their influence on accident frequency and severity. Field data collected from high-risk road sections, along with laboratory testing of pavement properties, provide a foundation for evaluating existing conditions. Using statistical and spatial analysis, the study identifies patterns in accident data, highlighting critical areas for improvement. Recommendations are developed based on these insights to propose cost-effective strategies that prioritize safety, such as optimized pavement materials, better signage, and improved road design. This research underscores the role of systematic traffic safety analysis in advancing safer, more resilient transportation systems. This project undertakes a comprehensive traffic safety analysis aimed at reducing accidents and improving road safety through a detailed examination of key factors affecting driving conditions. By focusing on parameters such as skid resistance, pavement surface texture, visibility, road geometry, and traffic flow, the study seeks to identify elements that significantly impact accident rates and driving safety.

Abstract-This study explores the potential of Kraft lignin and wood-based additives to mitigate rutting in porous asphalt concrete (PAC), a material widely used for its water permeability and noise-reducing properties. PAC, however, suffers from rutting, a type of pavement distress that leads to deformations and reduced performance under traffic loads. The research evaluates the impact of incorporating Kraft lignin and wood-based modifiers into PAC to enhance its rutting resistance. Experimental investigations, including wheel-tracking and Marshall stability tests, were conducted on asphalt samples with varying concentrations of these modifiers. Results indicated that both Kraft lignin and wood-based additives significantly improved rutting resistance, with lignin contributing to greater binder stiffness and wood additives enhancing aggregate bonding. These findings suggest that bio-based modifiers could offer a sustainable solution to improving the durability of porous asphalt pavements, reducing maintenance costs and environmental impact.

Abstract-Lifting bridges, also known as movable bridges, are crucial for enabling both road and maritime traffic, especially in regions where waterways intersect with busy transportation corridors. These bridges, including types such as bascule, swing, and vertical lift bridges, allow for efficient passage of vessels while maintaining road connectivity. Research into lifting bridges spans a range of disciplines, from structural engineering and materials science to automation and environmental impact studies. One primary focus is on the design and mechanics of movable bridges, with emphasis on the structural integrity, materials, and load-bearing capacities of these complex systems. Innovations in materials science have led to the exploration of corrosion-resistant alloys and high-performance composites, improving the durability and lifespan of lifting bridge components. Additionally, advanced automated control systems are becoming increasingly important, with research on robotic mechanisms and smart sensors aiming to streamline bridge operations and enhance safety. These innovations are complemented by studies into the impact of lifting bridges on traffic flow, which examine the operational challenges and disruptions posed by the periodic lifting and lowering of bridges. Another key area of research involves the environmental impact of lifting bridges. Studies have been conducted on the ecological effects of bridge operations on aquatic ecosystems, particularly in relation to waterway traffic and habitat disruption. Moreover, with the rise of sustainable infrastructure, researchers are exploring ways to reduce energy consumption and carbon footprints associated with the mechanical lifting process. Further, lifting bridges present unique challenges in extreme environments, such as those found in cold and hot climates, where materials and mechanisms face additional stresses due to thermal expansion, corrosion, or ice formation.

Abstract-The machine is a double-purpose unit proposed to chop and crush forage crops in an efficient way, to cut down on waste and inefficiency in agricultural practices. It discusses evaluation related to the performance of the machine, with emphasis on its productivity in trimming different forages. The study discusses the advantages the use of this machine would bring about, such as minimum labor costs and efficient crop management. Testing results show that the machine achieves the basic standards of operation for agricultural purposes. The main objective of the project was to develop a machine that efficiently performs chopping and crushing work simultaneously with the ability to overcome the weaknesses of machines that can only perform the two functions separately. This multi-purpose functionality aims at increased productivity and saving on operational costs. An increased need for environmentally friendly economical machines capable of delivering agricultural needs effectively, therefore, is essential to achieve economic sustainability.

Abstract-Chatbots function as software that enables users to ask questions and receive assistance through appropriate responses. This paper explores an AI-based chatbot designed to serve as an online ticketing system, streamlining the process of issue reporting, resolution, and user assistance across various domain. It also includes features like customer support, IT helpdesks, and event management. Natural language processing (NLP) is used by this proposed chatbot to understand user queries, categorize tickets, and provide instant responses. The aim of this chatbot is to enhance efficiency, reduce response times, and improve user satisfaction.

Abstract-Artificial Intelligence (AI) has become a pivotal technology for automating complex processes, while Soft Computing provides innovative ways to manage imprecise and uncertain data. By combining the two, hybrid systems leverage the strengths of AI’s precision and Soft Computing’s adaptability. This paper delves into the principles behind these hybrid models, emphasizing their use in healthcare, autonomous systems, finance, and smart cities. It also highlights the challenges of scalability and interpretability and outlines potential research directions, including integrating quantum computing and promoting explainable models.

Abstract-Low productivity of workers is the most significant factor behind delivery slippages in manufacturing industries. As manufacturing is a laborer predominant industrial sector, this paper focuses on worker output and their efficiency in the manufacturing sector. It covers the definitions of productivity, efficiency of the workers, its perspectives and the factors influencing the productivity. Proposed ANOVA method optimize performance of productivity and worker production parameters. Also observed more sensible case to increase production productivity.

Abstract-The challenges of urban areas with ever-increasing traffic congestion, emergency response, and maintaining road safety are the basis of this paper. The ITMS proposed in this paper treats optimization of timings at the traffic signals based on real-time vehicle counts, along with the detection of emergency vehicles and accidents, as its prime mandate. To achieve these objectives of optimal traffic management, advanced technologies, such as sensor detectors, algorithms for processing data, and communicating networks, were adopted. With simulations and evaluations, the ITMS holds great promise in enhancing traffic flow efficiency as well as reducing congestion while shortening emergency vehicle response times vis-a-vis fixed-time signal control. The research performed here addresses the development of more sustainable and resilient urban transportation systems.

Abstract-This study focuses on the design and analysis of a hollow shaft for an EV go-kart, optimizing weight reduction and structural integrity. Using SolidWorks for design and ANSYS for Finite Element Analysis (FEA), the shaft’s performance under mechanical stresses and cyclic loads was evaluated. Results demonstrated significant weight savings while maintaining strength, rigidity, and durability, enhancing the go-kart’s efficiency and reliability. This work highlights the potential of hollow shafts in improving EV performance through lightweight design.

Abstract-Employing generative adversarial networks, specifically with regard to cycle consistency loss and mask vectors, mainly concentrates on the colorization of Synthetic Aperture Radar (SAR). Most SAR imagery is devoid of chromatic information. Contemporary deep learning techniques are the predominant approach for SAR colorization. The methodology proposed herein employs a multidomain cycle-consistency generative adversarial network (MC-GAN). It enhances performance through the integration of a mask vector and cycle-consistency loss. The approach does not necessitate the availability of paired SAR-optical imagery. The multidomain classification loss contributes to the precision of the color output. The methodology has been evaluated using the SEN1-2 dataset for urban and terrain areas.

Abstract-The cost of commuting to and from school is a burden for many people, especially in urban areas. While ride-hailing services are popular worldwide, most students face issues with accessibility and convenience. The aim of this work is to create and use fairShare. A web platform that allows students to connect and share rides, thereby reducing transportation costs and reducing the environment around them. Users can register, post trips,and compete with other students using the same route. Early tests of the platform have shown that it reduces student travel costs and provides a good user experience. The platform also promotes sustainable practices for students. fairShare demonstrates the potential of student-friendly carsharing to reduce transportation costs and improve social interaction. The platform has the ability to measure a broader and more effective way for students to take action.

Abstract-With the advent of the “e-” revolution starting in 2000, the issue of cyber security, cyber-attacks and cyber threats which included domains, but not e-business, e-government, e-; commerce etc. only occurred because for the issue of cybersecurity in e- learning is under-explored, the aim of this paper is to present methods that focus on monitoring cybersecurity issues related to e- learning processes on. In addition, this article aims to present some good examples of cybersecurity management strategies in e- learning and cybersecurity trends in this area.[2] This paper will present possibilities for increasing information security and cyber- security awareness in education and e-learning that will inspire future cybersecurity professionals to navigate their career path.[3].

Abstract-The growing dependence on technology for a wide range of activities has dramatically increased computational demands, driving significant growth in computer network usage over the past few decades. This surge in demand for processing and storage capabilities has opened up business opportunities for companies but has also drawn the attention of cybercriminals. In response to these threats, researchers have developed various attack detection and prevention models. This paper introduces a new intrusion detection model that operates in two phases. The first phase involves building a feature ontology to train a convolutional neural network (CNN), and the second phase tests the trained model. For feature selection, the model uses an Elephant Herd Optimization-based genetic algorithm, which efficiently identifies a strong feature set for classifying network sessions. Experiments on a real-world dataset show that the proposed model can detect various types of attacks within normal sessions. Results demonstrate improved accuracy and performance metrics compared to existing models.

Abstract-IoT device-based communication boosts monitoring, business operations, and daily activities but also increases the load on servers and clouds. To handle this, edge computing acts as an intermediary layer. Efficient job management is critical for large-scale IoT networks, but existing models often fail to adapt based on past job sequences. This work introduces a model using a modified wolf Optimization algorithm to dynamically balance loads without prior training. It also incorporates a Random Forest model to generate initial job sequences. Experiments show that the proposed approach reduces job makespan time and enhances edge resource utilization compared to other models.

Abstract-Virtual meetings have become an essential mode of communication in contemporary professional environments. However, the fast-paced nature of virtual meetings undermines the ability to remember critical information accurately as even making notes is an imperfect mundane task, manual note-taking is both time- consuming and error-prone, often resulting in overlooked decisions and action items. SummrAIze is an AI-powered meeting assistant designed to address these challenges by automating the transcription, [1]summarization, and extraction of actionable insights during virtual meetings on platforms like Google Meet and Microsoft Teams. Using advanced machine learning algorithms, SummrAIze produces real-time summaries, highlights key points, and identifies action items, enabling participants to engage fully in discussions without sacrificing documentation accuracy. Integrated with productivity tools, SummrAIze not only reduces manual effort but also ensures that all essential information is recorded and accessible, enhancing team collaboration and workflow continuity. This paper presents the design, methodology, and potential impact of SummrAIze, a tool that redefines productivity in the context of virtual meetings.

Abstract-Non-destructive spectroscopic techniques represent the top-choice for any kind of process monitoring . Among all of the available techniques, Raman spectroscopy is one of the most solid and versatile tools to analyze several materials, both in lab and on-field conditions . Raman analysis has grown, reaching several industrial sectors such the food and textiles sectors .Raman spectroscopy displays several advantageous features over other techniques like infrared spectroscopy. For example, the quality of the signal collected is barely affected by the presence of water, allowing for use in plenty of applications where infrared analyses are not reliable . A representative case study is the in-situ monitoring of a fermentative process where Raman techniques outperformed any other spectroscopic approach .Molecular-level tissue characterization is highly potent for cancer diagnosis. As a tissue starts becoming cancerous, specific biomolecules are overexpressed or aberrantly expressed, which can be used as cancer molecular markers. If we can detect these molecular markers spectroscopically, it would lead to a new molecular-level cancer diagnosis with high objectivity.

Abstract-The plight of homeless children remains one of the most urgent global challenges, with millions of vulnerable children deprived of basic human rights such as shelter, healthcare, and education. Despite the rapid advancement of technology, child welfare systems in many developing countries still face significant hurdles, marked by inefficiencies and fragmented services. This paper proposes an innovative AI-driven system for adoption and rehabilitation that aims to address these systemic challenges holistically. By harnessing cutting-edge artificial intelligence (AI) algorithms, the system streamlines the adoption process, delivers personalized healthcare recommendations, and optimizes resource allocation for child welfare organizations. Through the integration of predictive analytics, data-driven decision-making, and a robust ethical framework, the system ensures transparency, fairness, and scalability. Early simulations and case studies highlight the transformative potential of AI in enhancing adoption success rates and improving healthcare outcomes for homeless children. The findings emphasize the system’s ability to drive meaningful improvements in global child welfare efforts, offering a scalable, ethical solution that can have a lasting impact on vulnerable children worldwide.

Abstract-Phishing attacks remain a prevalent cybersecurity threat, exploiting vulnerabilities in digital platforms to compromise sensitive user data. This paper introduces a novel machine learning-based framework for phishing URL detection, combining advanced feature engineering techniques and classification algorithms. By integrating lexical attributes, WHOIS data, and ranking metrics like PageRank and Alexa Rank, our approach enhances detection accuracy and minimizes false positives. Experimental results demonstrate superior performance across classifiers, achieving an accuracy of 99.8% using Support Vector Machines. The framework’s modular design ensures adaptability to evolving phishing tactics and scalability for enterprise deployment. This research lays the foundation for future advancements in AI-driven cybersecurity solutions.

Abstract-The demand for robust authentication systems has risen significantly as cyberattacks become increasingly sophisticated. Current authentication mechanisms, such as textual passwords, biometrics, and graphical systems, each have unique vulnerabilities. This research explores the concept of a 3D password system, which integrates various authentication schemes into a virtual 3D environment to enhance security. The system allows users to interact with objects in a 3D space, forming unique and complex passwords based on sequences of interactions. This paper elaborates on the system’s design, implementation, and potential applications in critical and non-critical systems. Detailed analyses reveal that the 3D password provides superior resistance to timing attacks, brute force attempts, and well-studied schemes, while maintaining user-friendliness. Future research avenues include the incorporation of AR/VR and IoT technologies to further expand the utility of the 3D password system.

Abstract-This paper proposes a flower recognition system using transfer learning with the ResNet-50 architecture. By utilizing pre-trained weights from ResNet-50, the system classifies ten species of flowers, drawing on an extended dataset with over 8,000 labelled images. The study addresses challenges in deep convolutional neural networks, such as overfitting and local optimality, by fine-tuning the ResNet-50 model. Initially, only the final layers of the model are retrained on the flower dataset, while the pre-trained layers remain frozen. After achieving initial convergence, all layers are unfrozen for full model fine-tuning. The dataset is divided into training, validation, and test sets to evaluate the model’s performance, which is measured using accuracy, and F1-score. The experimental results demonstrate that the transfer learning approach significantly improves classification accuracy and generalization, outperforming traditional methods. This approach proves especially effective in handling visually similar flower species and diverse environmental conditions. The study highlights the potential of transfer learning in enhancing the efficiency and robustness of flower recognition systems, contributing to broader applications in image classification tasks.

Abstract-The concept of Shoe Theory emphasizes that everyone is comfortable in their own shoes and should not be forced to wear someone else’s shoes. This theory posits that individual differences, including the effects of various elements such as time and generational perspectives, significantly impact workplace dynamics and organizational effectiveness. The theory highlights the importance of recognizing the unique experiences and backgrounds of team members to foster an inclusive and productive environment. Keeping creative destruction in mind, everything has its loophole to be breached. Although the answer may be yes or no, there always exists a condition of if/situation and but/exception.

Abstract-The k-Nearest Neighbors (k-NN) algorithm is a widely used non-parametric method for classification tasks, where the selection of the optimal value of k (the number of neighbors) plays a critical role in model performance. Traditional methods for selecting k, such as cross-validation or heuristic approaches, can be time-consuming and computationally expensive. This paper proposes an alternative approach to determining the optimal k for k-NN using polynomial regression. By treating the relationship between the value of k and the performance metric (such as classification accuracy) as a continuous function, we use polynomial regression to model this relationship and identify the k that results in the best performance. The polynomial regression model is trained on a set of performance data for different values of k, allowing for a smooth and accurate estimation of the optimal k across various datasets. Our experimental results demonstrate that the polynomial regression-based approach provides an efficient and effective method for selecting k, outperforming traditional techniques and reducing the computational cost associated with hyperparameter tuning. The proposed method also offers several advantages over traditional hyperparameter optimization techniques. By modelling the performance of k-NN as a continuous function of k, polynomial regression avoids the need for exhaustive grid search or cross-validation, making it particularly suitable for scenarios where computational resources are limited or time is constrained. Furthermore, the flexibility of polynomial regression allows for capturing complex, non-linear relationships between k and model performance, which can lead to more accurate predictions of the optimal value. Our approach is demonstrated one dataset, where it not only achieves higher accuracy but also reduces the overall time spent on model selection, making it a practical and scalable solution for hyperparameter tuning in machine learning applications.

Abstract-This paper reviews current research on Alumni Portal, the connections between alumnus and students, college interaction between alumnus, past records, event updates and records. The review covers 30 research papers, investigating database of Alumnus, students, past and present events held, interaction of alumnus in college events, interaction of alumnus and students. For improving the previous Alumni portals and projects related to Alumni.

Abstract-This paper explores the transformative potential of an Augmented Reality (AR) storytelling application designed to enhance traditional storytelling methods by integrating interactive digital animations, text, and audio into physical environments. The app offers a dynamic and immersive experience, particularly for children, by enabling real-time interaction with animated characters, voice narration, and engaging, interactive scenes. Unlike static books or conventional digital content, this app allows users to actively participate in the narrative, creating a more engaging and educational experience. By overlaying digital elements onto the real world, the app fosters increased interactivity and encourages deeper emotional and cognitive engagement with the story. Children can interact with animated characters, explore rich 3D environments, and receive instant feedback through audio cues and animations that respond to their actions. The app also supports educational growth by offering interactive learning modules, promoting reading comprehension, and allowing customization of story elements to accommodate multiple learning styles. The application leverages cutting-edge AR technologies to transform traditional narratives into immersive experiences, providing both entertainment and educational value. By integrating AI-driven components for voice recognition and dynamic content generation, the app can offer personalized experiences and adaptable content based on user preferences and interactions. This survey examines the underlying technologies and design choices that contribute to the app’s ability to engage users, as well as the broader implications of AR in storytelling for enhancing educational tools and creative learning platforms.

Abstract-This paper dives into how robotics is transforming manufacturing today. It looks at how robots are making processes faster, safer, and more efficient while also tackling some challenges like high costs and technical complexity. By exploring industries like automotive and consumer goods, and through examples from companies like Toyota and Unilever, the paper highlights both the advantages and limitations of using robots. It also touches on important issues like job impacts and cybersecurity risks, suggesting that thoughtful planning is essential for making the most of robotics in manufacturing.

Abstract-This paper examines the pivotal role of mechanical engineering in advancing transportation through innovations like electric vehicles, lightweight materials, and dual-fuel systems. It highlights their impact on sustainability, efficiency, and safety while addressing challenges such as costs, regulations, and public acceptance. Emerging technologies like Hyperloop and hydrogen propulsion are also explored, emphasizing their potential to redefine global mobility.

Abstract-Diabetes is one of the most frequent diseases worldwide where yet no remedy is discovered for it. Every year a great deal of money has to be spent for caring for patients with diabetes. Therefore, it is crucial that prediction should be very accurate and a very dependable method must be adopted for doing so. One of these methods is the use of artificial intelligence systems, and in particular, the use of Artificial Neural Networks, or ANN. So, in this paper, we used artificial neural networks in order to predict whether or not a person has diabetes. The criterion was to minimize the error function in neural network training with the help of a neural network model. After training the ANN model, the average error function of the neural network was equal to 0.01 and the accuracy of the prediction of whether a person is diabetics or not was 70%

Abstract-The enhancement in web technologies has contributed to the evolution of web applications that are very dynamic and engaging. This research work focuses on the creation of an online image editing application that is based on cloud infrastructure and modern web layouts/development tools such as Next.js, TailwindCSS, and Cloudinary’s APIs, among other resources, to deliver advanced image editing features. The application incorporates Clerk to allow users to create login accounts and easily register, while data is managed using MongoDB to facilitate the security of users and edited pictures across several devices. Necessary and basic features such as object removal, editing backgrounds, recoloring pictures, restoring, and changing the size of images are handled within the cloud and therefore benefit the functionality of the application and users as well. In addition, a contact form utilizing EmailJS has been integrated to enable communication with users. This research work highlights the legitimacy of cloud-based solutions as well as their expanded geographic reach in catering to an advanced user experience within image editing applications, thus supporting the growth of cloud computing and web technology.

Abstract-Due to the large amount of information generated every day, automatic writing is an important part of knowledge management. The discipline has made great progress, especially with the emergence of abstraction, abstraction and hybrid content models. In the extraction method, the main idea is preserved by selecting the main sentence or phrase from the text, while in the abstraction method, all the information is repeated to create new sentences. As the name suggests, hybrid models include the features of both extraction and abstraction systems to get the best of both approaches. However, issues remain, particularly in how to address the authenticity, coherence, and length of the text. This article examines the current state of writing concepts and topics in practice and future research.

Abstract-This study introduces the Car Surveillance System (Driver Negligence and Dissuader System), integrating advanced lane detection, drowsiness detection, pedestrian detection, and object detection technologies to boost road safety. Much like the luggage storage website, it presents a user-friendly interface and real-time alerts to avert accidents. Intelligent functionalities ensure efficacy and security, simplifying driving experiences and encouraging hassle-free travel. Tailored settings and transparent pricing cater to individual driver requirements, tackling prevalent challenges and nurturing safer roads for all users.

Abstract-In light of the increasing gun violence incidents worldwide, there is a pressing need for automated visual surveillance systems capable of detecting handguns. This paper presents a method for real-time handgun detection in video streams using the YOLO algorithm, comparing its performance in terms of false positives and false negatives against the Faster CNN algorithm. To enhance detection accuracy, we compiled a custom dataset featuring handguns from various angles and merged it with the Roboflow dataset. The YOLO model was trained on this combined dataset and validated using four different videos. The results indicate that YOLO effectively detects handguns across diverse scenes, demonstrating superior speed and comparable accuracy to Faster CNN, making it suitable for real-time applications.

Abstract-The field of healthcare is turning more towards tools to improve access, to services and make the experience better for patients and providers alike. A specific example is “Appointify,” a web platform for booking doctor appointments that was created using the MERN technology stack— MongoDB, Express.js, React and Node.js—with a goal of simplifying the appointment process and connecting patients, with healthcare professionals seamlessly. This document provides an outline of “Appointify ” a system created to tackle the issues encountered in appointment handling like extended waiting periods and disorganized scheduling well as the absence of efficient communication, between patients and healthcare providers.”Appointify” allows patients to search for doctors based on their expertise area request appointments access their history and update their profiles. It also equips doctors with functions to control their availability, schedule appointments. Engage with patients effectively. The platform includes functions such, as role based access control for security measures and encryption to safeguard data privacy It also features responsive design for user friendly interaction, on various devices

Abstract-In this research paper we have proposed a device that helps visually impaired people recognise currency denomination in order to detect the denomination of Indian currency. The members of this community have challenges particular to them when it comes to dealing with money, and as such there is an ever-growing need for quick and accurate identification tools appropriate for their scenario. We describe the process we have followed to develop the device, offering a blend of image processing and machine learning to allow currency identification in real time. Surveys of potential users revealed important preferences and needs for accessibility and ease of use, guiding the design of a new device system. According to test results, the device achieves high accuracy in denominations recognition and effective user satisfaction, demonstrating a potential device providing financially independent life for visually impaired users. These findings underscore the value of blending cutting-edge technology with user-centered design to create impactful solutions for underserved communities. The paper hence concludes with recommendations for the further enhancements and future research to expand the device’s features and accessibility.

Abstract-This system is designed to solve a common problem: running out of gas without knowing when it’s about to happen. The system keeps track of how much gas is left in the container by continuously checking its weight. If the gas is running low, it can automatically place a new gas order using the Internet of Things (IoT) technology. A device called a load cell is used to measure the weight of the gas container, and this data is sent to an Arduino Uno (a small computer) to compare with a standard weight. If the gas is low, the system sends a message to the user via SMS, using a GSM modem. For safety, the system also has sensors to detect gas leaks (MQ-2 sensor) and monitor the surrounding temperature (LM35 sensor). If any unusual changes are detected by these sensors, such as a gas leak or a sudden change in temperature, a siren will sound to alert the user.

Abstract-Liver diseases like cirrhosis and hepatitis are major causes of global morbidity and mortality, highlighting the need for early detection. Traditional diagnostic methods often identify liver damage at later stages, limiting preventive interventions. This study develops a machine learning model to predict liver damage earlier using clinical features and lab results. By analyzing a data-set with patient demographics and biochemical markers, we apply machine learning algorithms, including Random Forest, Decision Tree, and Logistic Regression, and evaluate their performance using metrics like accuracy, precision, recall, F1 score, and ROC-AUC. The Random Forest model outperformed others, showing high accuracy and robustness. Feature importance analysis revealed critical clinical factors, such as serum bilirubin and liver enzymes, in predicting liver damage. These results suggest that machine learning, especially Random Forest, could aid in the early detection of liver disease, improving patient outcomes. Future work will focus on using larger, more diverse data-sets and advanced models to improve predictive accuracy.

Abstract-Financial systems have become more complicated than ever before due to their fast growth, which calls for creative methods of managing, analyzing, and forecasting system behavior. In order to solve problems in intricate financial systems, this study investigates the use of intelligent computing and trustworthy machine learning models. The goal of the project is to improve decision-making, risk assessment, and anomaly detection in dynamic financial contexts by fusing cutting-edge computational techniques with reliable AI frameworks. The dependability and interpretability of machine learning models are given special attention in order to make sure they satisfy the exacting standards of accuracy and transparency that are necessary for financial stakeholders. The implications of these technologies for reducing systemic risks and enhancing operational effectiveness are also covered in the study. This study demonstrates the revolutionary potential of intelligent computing and reliable machine learning in creating robust and flexible financial ecosystems via case studies and experimental validations. The results highlight how important they are in determining how finance and economic stability develop in the future.

Abstract-For more effective treatment, early diagnosis of liver disease is crucial. Detecting liver disease in its early stages is challenging due to its subtle symptoms, often becoming apparent only in advanced stages. This research leverages machine learning techniques to address this issue by enhancing liver disease detection. The primary objective is to differentiate between liver patients and healthy individuals using classification algorithms. Liver disease has seen a global increase in prevalence in the 21st century, with nearly 2 million annual deaths attributed to it according to recent surveys. It accounts for 3.5% of global deaths [1]. Early diagnosis and treatment can significantly improve outcomes for patients with chronic liver disease, which is among the most fatal illnesses. The advancement of artificial intelligence, including various machine learning algorithms like Regression, Support vector machine, KNN, and Random Forest, offers the potential to extend the lifespan of individuals with Chronic Liver Disease (CLD).

Abstract-Concurrency and Synchronization in digital electronics where algorithms are use to comprehend the all the calculations for work. Digital machines ranging from Embedded Systems, IOT, Computers, Smartphones, Servers and Networking systems. Synchronization has became a very crucial part of basic programs running in the background of any operating system, that is the “Kernel”. These algorithms are the basic part of the OS for its smooth working in multi-programming, load balancing, time synchronization, data I/O ops within and out of the system, parallel computing with GPUs, I/O ops with IOT and cloud systems, Network and data security, mathematical calculations, etc. Synchronization programs are used to prevent conditions such as data races, deadlock, network latency, data corruption, manipulation and many more. Conditions created by these bugs can be visible or invisible in the user space. This Research paper is a comprehensive analysis on Concurrency and Synchronization. Source code examples of such conditions are given below from the original source code of some of the common linux distros. Applications of solutions to some of these issues in programs and systems to help progress for development of the performance and results.

Abstract-Dynamic Ride Pricing is a vital feature in the ridesharing industry that allows companies to adjust ride fares based on shifts in supply, demand, weather conditions, and other relevant factors. This study details the development of a machine learning-driven dynamic pricing model designed to optimize fare adjustments in real time. By analyzing key variables such as trip distance, weather, and historical patterns of supply and demand, the algorithm can deliver pricing that is both contextually relevant and responsive. The model aims to achieve a balance between profitability and customer satisfaction by swiftly adapting to fluctuating market conditions. Leveraging advanced machine learning techniques, it ensures pricing that is not only accurate but also fair and responsive. By integrating these factors into a unified pricing strategy, the model provides an optimized solution that enhances operational efficiency and meets consumer needs, ultimately contributing to a more equitable and efficient pricing system in the ridesharing sector.

Abstract-The use of wind power and conversion into energy, methodology regarding implementation of the idea, Advantages and Disadvantages and the scope for future.

Abstract-Augmented Reality, also known as AR technology, is a tool that employs computer graphics to superimpose a different layer of information onto the real world. Traditionally, virtual reality provided more interactive experiences when compared with other methods. In this paper, we explore the current state and future prospects of AR with a focus on its application in sectors such as medicine, education and retail among others. The functioning mechanisms of AR systems; sensors involved, processing algorithms required, rendering techniques for visual output and user interaction are discussed along with recent innovations like improved AR hardware or mobile applications. A literature review has been done to illustrate how AR enhances engagement in education, assists surgeons enhance precision during operations, changes customer experience in retail shops and provides entertainment through immersiveness. Moreover, AR technologies are also being explored for use in sectors such as tourism, automotive, and manufacturing, where they have the potential to revolutionize customer service, design processes, and workflow management.But there are obstacles that still hinders growth of AR such as technical barriers, privacy issues and expensiveness . Additionally, it discusses ways to overcome these challenges while pointing out things to research on so that maximum utility of AR can achieve. In conclusion, we find out that AR has great potential to alter different industries since it leads to more practical applications and encourages ongoing innovation.

Abstract-Chronic kidney disease (CKD) is a global health challenge, affecting millions of individuals and often leading to kidney failure when not detected early. The application of machine learning (ML) for CKD prediction has gained significant attention, enabling timely diagnosis using clinical data. This paper explores various ML techniques used for CKD prediction, focusing on preprocessing challenges such as missing data, data imbalance, and feature selection. Additionally, the paper discusses the emerging role of Federated Learning (FL), a decentralised approach to ML that allows for privacy-preserving collaborative model training across institutions.

Abstract-Machine learning techniques are doing wonders in every sphere of life but using predictive analysis in healthcare is a challenging task. However, if implemented properly these techniques help in making timely judgements about the health and treatment of patients. Globally, diseases including diabetes, heart disease, and breast cancer are major causes of death; yet, the majority of these deaths are due to failure to have regular checkups for these conditions. Low doctor-to-population ratios and a lack of medical infrastructure are the root causes of the above-mentioned issue. Thus, early detection and treatment of these diseases can save many lives. Machine Learning, Deep Learning and Streamlit is an effort concentrated on the development of healthcare using in-depth engines to forecast several sicknesses. Streamli Cloud and Streamlit Library facilitate deployment of prediction models like a breeze for developers. This has made accessing and using prediction capabilities of the system easily done by any layman. The paper focuses on forecasting three major diseases namely diabetes, heart failure and Parkinson’s disease by using an advanced ensemble of deep learning models as well as traditional machine learning techniques. Then again, merging Support Vector Machine (SVM) algorithm together with Logistic Regression models will form one such integration scheme.

Abstract-The goal of the web-based search engine “Intelli Search” is to give users accurate and pertinent content by combining personalized video recommendations with sophisticated AI-driven response production. The platform imitates Gemini’s capabilities by leveraging the YouTube API to suggest pertinent films arranged by comment engagement and the Gemini API to produce theoretical answers based on user inquiries. By using MongoDB to store and show user search history in a sidebar, the project allows users to view past queries after entering their login information. Auth0 securely manages authentication, guaranteeing a quick and secure user login. Through the integration of these technologies, Intelli Search provides a dynamic and customized user experience, enhancing search relevance by fusing multimedia resources with theoretical knowledge. The architecture is examined in this work.

Abstract-Medical image analysis is a critical component in modern healthcare, enabling more accurate and timely diagnoses. Deep learning techniques, particularly Convolutional Neural Networks (CNNs), have shown impressive capabilities in automating medical image interpretation. This paper reviews the latest advancements in deep learning methods for medical image analysis, covering key applications such as image classification, segmentation, and object detection. We discuss the challenges in applying deep learning models to medical imaging, such as the need for large annotated datasets, generalization to diverse datasets, and model interpretability. Additionally, we provide an overview of state-of-the-art architectures and their performance in different medical imaging tasks. Finally, we address the future directions and potential clinical applications of these techniques.

Abstract-Deep reinforcement learning has emerged as a transformative technology in AI and robotics, finding new answers to challenging problems in space exploration missions. This review details the latest developments within the DRL framework with applications in space robotics, exploring aspects such as autonomous navigation and resource optimization as well as mission planning. In this study, we do some case studies on strategies like AlphaNavNet, AstroPlannerNet, and open-source SpaceRL framework. We review how the DRL-based system addresses some key issues such as unpredictable terrain, delay in communication and exploration versus exploitation. In addition, this paper covers the embedding of simulation-to-reality translation in robotics and astrophysical modeling and the application of deep learning techniques such as Double Deep Q- Networks (DDQN) and Reinforced Deep Markov Models (RDMM) in augmenting the decision- making power of space missions. Although DRL has proved to outperform other approaches in simulaions and prototype testing, the review also emphasizes experimentation for added robustness and reliability within extraterrestrial condition. Through this analysis, we gain insight into the potential and limitations of DRL in advancing space exploration, using new architectures and real-world validation.

Abstract-AI has deeply penetrated even the most critical domains, including healthcare, finance, and governance, making it possible with its transformative potential to reach unprecedented efficiency and innovation. Still, this widespread diffusion poses ever more urgent challenges related to ethics and accountability that should not be ignored. Synthesizing insights from five seminal studies on “Ethical Approaches in Designing Autonomous and Intelligent Systems,” “Accountability of AI Under the Law: The Role of Explanation,” “Explainable AI as a Tool for Accountability,” “AI Accountability in Financial Decision-Making,” and “Ethical Implications of Artificial Intelligence (AI) Adoption in Financial Decision-Making,” this paper explores the interplay between accountability frameworks and explainable AI (XAI), regulatory compliance, and societal impacts by combining theoretical and practical perspectives. This paper explores the necessity of explainable models in terms of handling ethical dilemmas, such as bias mitigation, fairness, and transparency, through technical methodologies like sensitivity analysis, counterfactual reasoning, and Shapley values for feature importance. Case studies in health care, finance, and governance -AI-driven diagnostics, credit risk assessments, and algorithmic decision-making in welfare systems- will be explored to illustrate consequences of opacity and betterment facilitated by accountability-driven approaches. In terms of these elements, this paper discusses emerging regulatory landscapes, including the AI Act in the European Union and global data protection laws, as importance factors forming the ethical practices of AI. Public trust erosion due to biased or opaque AI systems is a further societal impact, and inclusive design and multi-stakeholder accountability are put forward as important aspects in this context. A balanced framework of ethical considerations to guide AI innovation should encompass both technical and normative dimensions. Various practical recommendations are laid out, such as standardized practices of XAI, robust accountability mechanisms, and proactive approaches to compliance and regulatory matters. The research brings the technological advancement closer to the imperatives of ethics in AI, toward trust, equity, and justice in its use.

Abstract-The use of DL algorithms revolutionizes the approach towards the detection of plant disease, making this most critical agricultural technology develop towards accuracy and efficiency that were not possible even with earlier methods. Apart from the benefits that an automated system may have over a manual intervention one, such as quicker identification of disease and less manual efforts, DL techniques, and CNNs in particular, allow the diagnosis of the diseases on plants with precision. The potential of AI-powered systems for plant disease detection is the ability to automatically analyze a plant image to recognize the symptoms and classify diseases with high accuracy. These systems also have the potential to provide real-time support by analyzing complex images and suggesting management recommendations for diseases. Thus, with DL algorithms, the system can identify diseases in plants, detect slight changes in texture and color, and recommend the corrective action to optimize crop health. Further, with the recent advancement in optimized models like YOLOv5 and hybrid techniques by integrating CNN with traditional classifiers such as Support Vector Machines (SVMs), the accuracy in detection has increased. Although the approaches present promising outcomes, challenges abound, especially in dealing with complex image backgrounds, low-quality datasets, and computational efficiency. This paper discusses approaches designed to overcome these hurdles, thus indicating the future direction of plant disease detection systems. This work will, therefore contribute towards the advancement of AI-driven agricultural solutions in terms of the accuracy and speed of plant disease detection and enable better crop management practices around the world.

Abstract-With the increasing prevalence of limited labelled data in many real-world applications, few-shot learning (FSL) has become an essential approach to enable effective learning from minimal examples. However, scalability, domain generalization, and adaptability to new tasks remain significant challenges. This paper introduces “Unified Adaptive Few-Shot Learning”, a novel framework that combines the strengths of metric learning, graph neural networks (GNNs), and meta-learning. By extending Prototypical Networks with GNN- based prototype refinement, our approach improves the quality of class representations and captures complex inter-class relationships. Meta-learning further enhances task-specific adaptation, while self-supervised pretraining boosts feature robustness. Additionally, integrating class metadata facilitates seamless transitions between few-shot and zero-shot tasks. Experimental evaluations on benchmark datasets like Mini-ImageNet and Meta-Dataset demonstrate that our framework outperforms existing methods in accuracy, scalability, and cross-domain generalization, offering a promising solution for real-world FSL applications.

Abstract-The manufacture of raw materials to deliver the product to the consumer in a traditional supply chain system is a manual process with insufficient data and transaction security. It also takes a significant amount of time, making the entire procedure lengthy. Overall, the undivided process is ineffective and untrustworthy for consumers. If blockchain and smart contract technologies are integrated into traditional supply chain management systems, data security, authenticity, time management, and transaction processes will all be significantly improved. Blockchain is a revolutionary, decentralized technology that protects data from unauthorized access. The entire supply chain management (SCM) will be satisfied with the consumer once smart contracts are implemented. The plan becomes more trustworthy when the mediator is contracted, which is doable in these ways. The tags employed in the conventional SCM process are costly and have limited possibilities. As a result, it is difficult to maintain product secrecy and accountability in the SCM scheme. It is also a common target for wireless attacks (reply to attacks, eavesdropping, etc.). In SCM, the phrase “product confidentiality” is very significant. It means that only those who have been validated have acc ess to the information. This paper emphasizes reducing the involvement of third parties in the supply chain system and improving data security. Traditional supply chain management systems have a number of significant flaws. Lack of traceability, difficulty maintaining product safety and quality, failure to monitor and control inventory in warehouses and shops, rising supply chain expenses, and so on, are some of them. The focus of this paper is on minimizing third-party participation in the supply chain system and enhancing data security. This improves accessibility, efficiency, and timeliness throughout the whole process. The primary advantage is that individuals will feel safer throughout the payment process. However, in this study, a peer-to-peer encrypted system was utilized in conjunction with a smart contract. Additionally, there are a few other features. Because this document makes use of an immutable ledger, the hacker will be unable to get access to it. Even if they get access to the system, they will be unable to modify any data. If the goods are defective, the transaction will be halted, and the customer will be reimbursed, with the seller receiving the merchandise. By using cryptographic methods, transaction security will be a feasible alternative for recasting these issues. Finally, this paper will demonstrate how to maintain the method with the maximum level of safety, transparency, and efficiency.

Abstract-Cross-site scripting is one of the severe problems in Web Applications. With more connected devices which uses different Web Applications for every job, the risk of XSS attacks is increasing. In Web applications, hacker steals victims session details or other important information by exploiting XSS vulnerabilities. We studied 412 research papers on cross-site scripting, which are published in between 2002 to 2019. Most of the existing XSS prevention methods are Dynamic analysis, Static analysis, Proxy based method, Filter based method etc. We categorized existing methods and discussed solutions presented on papers and discussed impact of XSS attacks, different defensive methods and research trends in XSS attacks.

Abstract-The Focus Pro Anti-Distraction Application is a productivity-enhancing tool designed to help users maintain focus by reducing distractions from digital platforms like social media, videos, and other time-wasting activities. With the increasing prevalence of digital distractions, this app provides a structured, customizable solution to improve concentration and task completion for students, professionals, and anyone seeking better focus. The app offers multiple focus modes, each tailored for specific tasks: Learning Mode, Assignment Mode, and Notes Mode. These modes feature task management tools, reminders, progress tracking, and a calendar to organize tasks and goals effectively. Users can customize their experience based on their specific needs, whether they are studying, working on assignments, or taking notes. A standout feature is the app’s blocking functionality, which allows users to create a customized list of websites and apps to block during use. This helps users avoid distractions and stay on task by preventing access to non-productive content on both mobile and desktop devices. In addition, the app integrates an AI-powered Filtering system that intelligently analyzes content on platforms like YouTube and Google. It uses keyword and hashtag analysis to allow access only to study-related content, ensuring users remain focused on educational materials. The app also includes performance analytics, which tracks user productivity and provides insights into task completion. Users earn points for completing tasks on time, and these points contribute to earning badges. This gamification approach encourages users to stay motivated and improve their focus. In addition, the app offers a streamlined profile section that allows users to monitor their achievements, track badges earned. The interface is designed to be user-friendly and visually engaging, making it easy for users to navigate modes.

Abstract-Within the face of world populace increase, sustainable and efficient crop production has come to be important. the mixing of emerging technologies consisting of the net of things (IoT), cloud computing, and machine mastering is revolutionizing agriculture through permitting actual-time soil tracking, crop selection, and predictive analytics for more desirable choice- making. This paper offers a comprehensive framework for IoT-enabled precision agriculture, which employs numerous sensors to reveal soil parameters—including moisture, pH, and temperature—and leverages advanced machine learning algorithms for crop advice and soil nutrient management. The proposed structures now not best optimize irrigation and fertilization but additionally provide a low-value, electricity-efficient method to information collection via wi-fi sensor networks. additionally, cloud-primarily based structures and cell programs provide farmers with far flung get entry to real-time data, permitting well timed interventions. by way of combining reinforcement learning fashions, multi-sensor information fusion, and modular hardware setups, this machine supports sustainable farming practices and will increase crop productiveness. The consequences show sizeable upgrades in prediction accuracy, decreased environmental effect, and more advantageous selection-making skills for farmers, contributing to the modernization of agriculture.

Abstract-The rapid advancements in deep learning (DL) techniques have transformed the healthcare sector, leading to notable improvements in diagnostic accuracy, personalized treatment, and ongoing patient monitoring. One particularly promising application of deep learning in healthcare is Human Activity Recognition (HAR), which uses wearable and mobile sensors to track and categorize individuals’ daily activities. HAR, especially within the framework of the Internet of Healthcare Things (IoHT), has demonstrated significant potential in enhancing elder care, rehabilitation processes, and chronic disease management. However, despite these advancements, several challenges persist in fully leveraging deep learning for healthcare applications. A major challenge is the dependence on large, labeled datasets for training models. In real-world scenarios, obtaining labeled data for HAR tasks can be time-consuming, costly, and often impractical, leading to a reliance on weakly labeled or unlabeled data. To tackle this issue, recent strategies in deep learning, particularly semi-supervised and reinforcement learning techniques, have been introduced to make efficient use of the vast amounts of unlabeled data available. These methods, such as Deep Q-Networks (DQN) and auto-labeling schemes, significantly lessen the manual labeling burden while preserving high model accuracy. Additionally, deep learning’s capability to integrate multi-modal data from various sensors (like accelerometers, gyroscopes, and context sensors) is vital for HAR tasks. This integration of sensor data offers a more thorough understanding of human activity and improves the accuracy of activity classification models. Among the most promising deep learning models for HAR are Long Short-Term Memory (LSTM) networks, which excel at processing sequential data typical in human activity monitoring. LSTMs effectively capture temporal dependencies in sensor data, making them well-suited for identifying complex motion patterns and contextual changes.

Abstract-The COVID-19 pandemic has had an unprecedented impact on health systems in most countries, and in particular, on the mental health and well-being of health workers on the frontlines of pandemic response efforts. The purpose of this study is to provide an evidence-based overview of the adverse mental health impacts on healthcare workers during times of crisis and other challenging working conditions and to highlight the importance of prioritizing and protecting the mental health and well-being of the healthcare workforce, particularly in the context of the emotional intelligence.

Abstract-Phishing is a category of cyberattack that includes the theft of credit card numbers, passwords, and other private data. We have employed machine learning algorithms to identify phishing websites in order to prevent phishing fraud. The availability of several services, including social networking, software downloads, online banking, entertainment, and education, has sped up the development of the Web in recent years. Consequently, enormous volumes of data are downloaded and uploaded to the Internet on a regular basis. Attackers can now obtain private information, including social security numbers, account numbers, passwords, and usernames, as well as financial information. This is one of the most important problems with web security and is referred to as a “phishing” attack on the internet. To identify these malicious websites, we employ a variety of machine learning methods, including KNN, Naive Bayes, Gradient Boosting, and Decision Trees. The study is broken down into the following sections. The introduction outlines the tools, methods, and concentrated zones that are employed. The process of gathering the data needed to proceed is described in depth in the preliminary section. Subsequently, the paper highlights the thorough examination of the information sources.

Abstract-Emerging topics such as environmental protection and energy utilization have pushed research and development of electric vehicles. In the last few decades, numerous technologies have been developed for EV importance. In this article, key research topics in the area of EVs, namely electric machines, electrochemical energy sources, wireless charging infrastructure, and latest EV/HEV models are covered. This Review paper aims to consolidate the key emerging technologies in this field and provide the readers a blueprint to begin their own journeys.

Abstract-This project introduces the YouTube Video Transcribe Summarizer, a tool designed to automatically extract transcripts and generate concise summaries from YouTube videos. By leveraging the YouTube Transcript API, the system retrieves accurate video transcripts and utilizes Google Gemini Pro’s advanced text-based model to create coherent summaries.
Users can input a YouTube video URL, which displays the video thumbnail for context. The application features a customizable prompt template to tailor the summary generation process, ensuring relevance to individual needs. Built on a user-friendly Streamlit interface, this tool aims to enhance content accessibility and engagement. Additionally, the project explores the possibility of executing local models for improved performance and user control. By streamlining the summarization of video content, the YouTube Video Transcribe Summarizer facilitates more efficient information consumption, empowering users to navigate the vast landscape of online video more effectively.

DOI: 10.61137/ijsret.vol.10.issue6.370

Indian Man Made Islands Idea to Save Wildlife/strong>
Authors:-Deepak Singh