VOLUME 13, ISSUE 4, APRIL 2024

Abstract: In this paper, we discuss a practical mechanism of digital matched filtering  which maximizes of output SNR for square, triangular , Gaussian pulse signals and other pulse signals in presence of additive white Gaussian noise (AWGN) by using a digital matched filter (DMF) corresponding to time domain convolution algorithm of input and reference signals using Cyclone II EP2C70F896C6 FPGA from ALTERA placed on education and development board DE2-70 with the following parameters: sampling frequency , pulse width , pulse period , samples number (length of reference signal) is 300, the ratios of signal to the noise at the input of the filter is , processing gain factor is 25dB. The results of filter operation are evaluated using a digital oscilloscope to display the input and output signals for different .

Keywords: DMF , Square Pulse , DPNG , DDFS , FPGA. Cite: Dr. Kamal Aboutabikh, Dr. Amer Garib,"Design and Implementation of a Digital Matched Filter for Square Pulses Signals using FPGA", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13401.

Abstract: The use of the internet across different generations has been on the rise recently. By providing unmatched connectivity, convenience, and access to other sets of information, the internet has also had severe effects, and phishing is one of them. Recent studies have also shown that the attacks on the aging population have been rampant, and they are expected to increase as cyber criminals become more complicated and ruthless in the execution of their strategy. While phishing attacks agonist multiple factors have caused older people, this study focused on the relationship between age and susceptibility to pushing attacks. A sample size of 140 participants was asked the same question, and results were calculated from a scale of 0-10. The main findings of this analysis were that only 2.1% of the variation in susceptibility can be explained by the use of the variation in age; the correlation coefficient of 0.1448 indicates a low but positive relationship between age and susceptibility to phi. The p-value of 0.6871 is more than 0.05 level of significance, which means that age is not a significant predictor of susceptibility to phishing attacks. Cite: Tunbosun Oyewale Oladoyinbo,"IMPLICATIONS OF PHISHING SCAM ACTIVITIES IN ADULTS BETWEEN AGE 50-80 IN THE UNITED STATES OF AMERICA", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13402.

Abstract: This work introduces a comprehensive music recommendation system that harnesses the power of artificial intelligence (AI) to understand and cater to human emotions. The system begins by curating a vast dataset comprising songs annotated with emotional attributes, meticulously collected from various sources. Leveraging advanced machine learning techniques, including sentiment analysis and feature extraction from audio signals, the system trains models to discern the nuanced emotional dimensions embedded within music. Through an intuitive user interface, individuals interact by either expressing their current emotional state or selecting from predefined emotional categories. Subsequently, the system utilizes this input to generate tailored music recommendations, ensuring that the suggested tracks resonate harmoniously with the user's mood. An iterative feedback loop allows users to rate there commendations, fostering continuous refinement and improvement of the recommendation algorithms. The system& deployment as a user-friendly application empowers individuals to effortlessly discover music that not only entertains but also resonates deeply with their emotional landscape, enhancing their overall listening experience. This work represents a significant advancement in personalized music recommendation systems, bridging the gap between AI technology and human emotion in the realm of music discovery and enjoyment.

Keywords: Machine learning, Music recommendation, MIDI, Multimodal fusion, Feature Extraction, Filtering Techniques

Abstract: This document explores the current market trend around Artificial Intelligence (AI), highlights the differences between human intelligence and AI, highlights the current importance of AI, lists its benefits and discusses its impact on various aspects of our lives. It also includes a study that assesses the future impacts, opportunities and risks associated with AI technology. It addresses the challenges of the future due to the development of artificial intelligence and examines the impact of artificial intelligence on the future landscape. The technology of artificial intelligence has a rich history characterized by continuous development and growth. Artificial intelligence focuses at its core on the development of intelligent agents. These include devices that can sense their environment and take action to maximize the probability of achieving predetermined goals. In today's digital world, artificial intelligence enables machines, computer programs and systems to perform tasks that normally require human intelligence, such as problem solving, inference and decision making. In addition, most computer systems have learning capabilities that enable iterative improvements in performance over time. Recent advances in artificial intelligence tools, including machine learning, deep learning, and predictive analytics, aim to improve skills such as observation, learning. , reasoning, cognition and decision making. This article explores the basic principles of modern artificial intelligence and looks at various representative applications in various fields. Cite: M. Arif Arshad, Prof.Dr.R.Nagarajan, "A SURVEY ON CONCEPTS OF ARTIFICIAL INTELLIGENCE AND ITS FUTURE SCOPE", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13404.

Abstract: The project presents Wellness online, a ground-breaking online application based on technologies such as MongoDB, Firebase, MUI, Axios, JWT token, and React.js. This all-inclusive software offers administrators, physicians, and patients customized dashboards that completely rethink healthcare management. Administrators have access to powerful tools that allow them to verify user accounts, analyze system statistics, manage appointments and fees, and get patient feedback. Simplified processes for managing appointments, uploading prescriptions, and reviewing patient comments are advantageous to doctors. With features including prescription access, payment processing, appointment scheduling, and feedback provisioning, patients have a smooth experience. Wellness Web seeks to improve doctor-patient communication, streamline appointment scheduling, and give administrators useful information for streamlining processes. Cite: Dr. J. JEYABOOPATHIRAJA, AND KEERTHANAA SREE P, "WELLNESS WEB", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13405.

Abstract: The solar helmet with a cooling system is a groundbreaking project that combines renewable energy and user comfort. This innovative headgear incorporates photovoltaic cells to harness solar power, supplying energy to an integrated cooling system. The abstract highlights the synergy between sustainable technology and personal well-being, showcasing a solution that not only generates clean energy but also prioritizes user comfort through effective temperature regulation. An Arduino nano microcontroller processes data from a gyroscope sensor, enabling real-time tracking of head movements for enhanced user comfort and safety. The solar helmet project aims to integrate photovoltaic technology into a wearable, efficient, and aesthetically pleasing design. By harnessing solar energy through the helmet, it seeks to provide a sustainable power source for various electronic devices, enhancing user convenience and reducing dependence on traditional charging methods. The abstract encapsulates the innovative fusion of technology and fashion, promoting eco-friendly solutions in everyday life.

Keywords: solar panel, IOT (internet of things).

Abstract: One of the major obstacles that the Internet retail industry faces is the widespread availability of counterfeit goods. These false goods imitate the look of real branded products, which poses a serious problem for both the industry and consumers. It's shocking to learn that about 30% of the things sold online are fake. Blockchain technology has attracted more attention as a response to this expanding problem, providing creative ways to reduce the ubiquity of counterfeit goods. Supply chain authenticity and transparency are guaranteed by blockchain's decentralized and impenetrable structure. By offering a traceable and secure framework that protects customers from purchasing fake items, this technology has the potential to completely transform the online retail industry. This paper proposes a decentralized blockchain solution to empower consumers in identifying the originality of the products independently of distributors. By establishing a blockchain network with anti-counterfeiting features, manufacturers can deliver goods without relying on traditional outlets, reducing quality assurance costs. The system utilizes blockchain technology to securely store product details, enabling verification against genuine information to identify counterfeit items. The verification procedure is streamlined by a smart contract-driven method, which enables producers to register items with distinct digital identities. Customers may use QR codes to confirm the genuineness of products.

Keywords: Fraudulent goods, Quick Response (QR) code, Blockchain, smart contracts, decentralized. Cite: Dr.Umesh Akare, Prof. Girish Umaratkar, Parag Wadhai, Aayush Kharwade, Pranav Bante, Ayush Dandekar, "RECOGNITION OF FRAUDULENT PRODUCTS USING BLOCKCHAIN", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13407.

Abstract: In the twenty-first century, amidst an overwhelming amount of data, the task of finding personally relevant content, particularly in the realm of movies, has become increasingly difficult. To address this issue, movie recommendation systems have emerged as indispensable tools, with the goal of making it easier to choose movies from a large number of options. This paper proposes a content-based approach to movie recommendation that uses machine learning to analyze movie attributes like genres, directors, actors, and plot keywords. By parsing and transforming movie metadata into meaningful representations, our system aims to provide personalized movie recommendations based on individual preferences. Using a variety of datasets, including important metadata such as actors, directors, and genres, we use algorithms such as Text Vectorization and Cosine Similarity to generate recommendations based on each movie's unique characteristics. This content-based filtering approach provides users with a personalized and enriching movie selection experience, addressing the issue of choice overload in the media environment of today.

Keywords: Movie Recommendation System, Recommendation Systems, Content-Based Approaches, cosine similarity

Abstract: India, predominantly reliant on agriculture, suffers an estimated 18% loss in global crop yield annually due to pest attacks, amounting to approximately Rs. 90,000 million. Overuse of pesticides poses numerous hazards including soil degradation, acute toxicity to humans and animals, shifts in pest populations, high control costs, and environmental residue issues. Whiteflies are particularly problematic pests, infesting plant leaves, excreting sticky honeydew, causing leaf discoloration or death, and reducing crop yield. Traditionally, farmers have relied on visual assessments to gauge whitefly infestations, but this method is often imprecise due to varying identification skills and the time-consuming nature of laboratory inspections. Given the economic importance of crops and the severe impact of pest damage, early detection of whiteflies has become imperative. To address this, we propose an Android application that calculates the affected area of plants and determines disease severity. The application also provides treatment recommendations in Hindi for identified diseases. Detection of plant diseases is a critical research area, offering benefits in monitoring vast agricultural fields. Automated disease detection through image processing offers a more accurate and efficient alternative to manual visual identification, which is prone to errors and time constraints. This approach enhances accuracy and facilitates timely intervention and disease management, ultimately improving crop productivity and sustainability.

Keywords: : Image Processing, Plant Disease, HSV(Hue Saturation Value), Machine Learning. Android Application

Abstract: Fit-Finder is an innovative website that solves the challenge of finding the best gym in today's health-conscious age. Fit-Finder simplifies the exercise discovery process by providing a user-centric platform that makes it easy for users to find gyms, discover features, training plans, and interventions with experts online. Going beyond traditional methods of engaging with fitness centers online, Fit-Finder uses an advanced search engine to curate individual gym listings. The application contains comprehensive information about gyms and provides one-time service to users in the selected region. Fit-Finder's focus on user needs is reflected in features like a fitness launcher and detailed gym information to help you decide based on fitness goals. The platform allows direct communication with gyms, facilitating interaction where users can explore and compare different packages. Fit-Finder recognizes the difference between online exercise and includes a virtual training section to isolate problem areas. Fit-Finder leverages technologies such as Java, Kotlin, XML, cloud-based React, and powerful data management to create an environment that is not only technological but also practical across multiple devices and platforms. This commitment to technology enhances the user experience, making Fit-Finder a versatile and user-friendly solution in an increasingly digital world. Fit-Finder's holistic Fitness experience shows that her passion for health goes beyond physical activity. While Fit-Finder has a detailed list of gyms, it goes a step further and includes features like virtual training, recognizing growth in online fitness, and allowing users to participate in activities like Zumba, MMA, arm wrestling, and more. Fit-Finder also provides links to certified doctors and physical therapists, demonstrating its commitment to solving a variety of health problems and promoting the right path to health. Fit-Finder's comprehensiveness, user-friendliness and innovation make it stand out. Stay at the forefront of health and wellness and contribute to the growth of the global community that values and values health. In summary, Fit-Finder represents a revolution in the way people find and interact with fitness centers. Its user-centric approach, advanced technology integration, and commitment to providing a holistic fitness experience places it at the forefront of fitness research. As Fit-Finder continues to evolve, it has the potential to not only change the way people exercise, but also lead to a healthier, stronger world.

Keywords: Include at least 4 keywords or phrases.

Keywords: Web 3.0, Philanthropy, Digital era, Solidity, React frontend, Ethereum ecosystem, Transparency, Efficiency, Smart contracts, Decentralization, Charitable giving, Fundraising, User empowerment, Security measures, Blockchain technology, Cryptocurrency wallets, User-centric design, Continuous improvement, Community engagement, social impact Cite: Umesh Aakre, Aparna Bondade, Samyak Sukhdeve, Masoora Khan, Gopal Kharwade, Vishal Tarwatkar, "DECENTRALIZED PLATFORM FOR CHARITY & CROWD FUNDING", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13411.

Abstract: Not only can wounds impair a patient's physical and mental well-being, but they also result in significant medical expenses. In the meantime, there may be a physician scarcity in some locations, and clinical exams may not always be accurate in diagnosing wounds. Accurate wound analysis is crucial for its diagnosis, management, and care. Nowadays, machine learning has become the most widely used method for wound image interpretation due to its rapid development in the fields of computer vision and medical imaging. This work examines the state-of-the-art deep learning research on wound image processing, encompassing segmentation, detection, and classification. Firstly, we examine the pre-processing techniques utilized in wound image analysis and assess the publicly available datasets from different studies. Secondly, different models applied to diverse machine learning tasks (identification, classification, and segmentation) and their applications in different types of wounds (e.g., burns, cuts, lacerations) are investigated. In conclusion, we address the difficulties encountered using the field of machine learning wound image analysis and offer a future direction for research and growth.

Keywords: Wound diagnosis, Computer Vision, Detection, Segmentation. Cite: KB Mangala, Umema Zaib, Divya M, Sarvar Begum, "Algorithms for determining the Injuries: A Survey", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13412.

Abstract: Menstruation remains a major impediment to quality of life for many, as persistent stigma and inadequate resources magnify struggles managing periods. This paper proposes BlissfulCycle, an innovative website fostering holistic menstrual wellness through five key capabilities. Users can track cycles, access vetted education content, join anonymous forums to share experiences, schedule affordable telehealth consultations and conveniently order essential products. Further self-care resources including yoga videos, curated playlists, nutritious recipes, inspiring quotes and a restroom locator promote physical and emotional comfort. By providing an integrated hub consolidating knowledge, community support, personalized care and destigmatized dialogue, alongside enhanced convenience accessing supplies, BlissfulCycle aims to empower individuals to understand and navigate their menstrual health confidently. Ongoing user surveys will iteratively refine BlissfulCycle's features and user experience to ensure optimal utility.

Keywords: Menstrual health, women's health, web platform, self-care, community. Cite: Rakesh M R, Deeksha K, Greeshma C Poojary, Harsharaj B, Kiran Kumar V, "BlissfulCycle: An Innovative Website Fostering Holistic Menstrual Well-Being", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13413.

Abstract: The aim of this research project is to create a Morse code detecting system by utilizing the ESP32 microcontroller's capabilities and machine learning capacity. An enduring communication technique, Morse code can be used for emergency signals and low-power communication among other things. In order to develop a flexible and effective Morse code detector, we plan to integrate contemporary technology with conventional communication methods. The main goal of the project is to create a small, inexpensive device that can precisely recognize and decode inputs that contain Morse code messages. The core processing unit, the ESP32 microcontroller, is responsible for preprocessing and signal acquisition. It also offers smooth networking choices for remote control and data transfer. In order to recognize Morse code signals, we use machine learning methods.

Keywords: ESP32, Convolutional neural networks, Recurrent neural networks, Signal preprocessing, Audio capture, Real-time recognition, internet of Things (IoT), Remote monitoring, Communication technology, Emergency signaling, Low-power communication. Cite: Dr.U.P. Akare, Prof. Kalpana Bhure, Adarsh Sonkusre, Atharva Ganorkar, Mayank Barapatre, Pratyush Roychowdhury, "Morse Code Detector Using Machine Learning", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13414.

Abstract: Cloud computing has revolutionized the way we store, process, and access data. As one of the leading cloud providers, Amazon Web Services (AWS) has played a significant role in shaping the industry and driving innovation. However, as new technologies and trends emerge, the future of cloud computing is constantly evolving. This paper aims to explore the future of cloud computing with a specific focus on AWS. Through a literature review, survey analysis, and case studies, we examine the major changes and challenges facing the industry, the impact of artificial intelligence and machine learning, the potential of quantum computing, and the emergence of blockchain. Overall, this paper provides valuable insights into the future of cloud computing and offers guidance for organizations looking to leverage the power of the cloud to drive innovation and growth.

Keywords: Cloud Computing, Amazon Web Services (AWS), Future Trends, Data Storage, Data Processing, Innovation, Artificial Intelligence (AI), Machine Learning (ML), Quantum Computing ,Blockchain ,Industry Challenges, Organizational Growth, Technological Evolution. Cite: Lankoji Venkata Sambasivarao, Kattamuri Ganesh Kumar, Potluri Sarath Chandra, Nihit Surya Naga, Ishwarya Rani Galla, "AWS And Future of Cloud Computing", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13415.

Abstract: In this study, we dive into the amazing world of robotic arms and explore how they can effectively identify, track, and manipulate objects in dynamic environments. We start by discussing the fundamental components and workings of robotic arms. Then, we explore object detection and recognition techniques like computer vision and machine learning to help the robotic arm accurately identify and categorize objects. Planning and control are crucial aspects, where we delve into planning algorithms for determining optimal trajectories and movements, as well as control algorithms to ensure precise and efficient manipulation of objects. Sensors, such as cameras and force sensors, provide vital feedback about the environment, while actuators like motors and pneumatic systems enable the arm to physically manipulate objects. Throughout our research, we address challenges and discuss potential future directions in this exciting field. By improving the capabilities of robotic arms in object segregation, we hope to contribute to the development of intelligent and efficient robotic systems.

Keywords: Robotic Arm Technology, Object Detecting, Servo Motors, Sensors, Robotic Arm Control Systems.

Abstract: Safety, reliability and cost are the major driving factors for research in automotive applications. The current suspension systems available today prove to be a bit lackluster by compromising on either the ride quality or stability of the vehicle. By compromising stability, traction also gets compromised. Automotive suspension systems play a vital role in ensuring the comfort and safety of a vehicle. Our project proposes an adaptive version of the semi-active suspension system with a new design that focuses on cheaper production and better stability. It aims to tackle the above-mentioned issues by monitoring the conditions and determining the type of irregularity present in the road ahead, and reacting to these irregularities and conditions by pumping a specific amount of oil into the suspension system in order for the suspension to stiffen or soften accordingly. The suspension system proposed by us displays the integration of Artificial Intelligence (AI) and IOT together that aims at improving the traction and holding the position of the vehicle to the surface of the road thus improving the stability of the vehicle, reducing body roll and trying to prevent motion sickness, leading to better riding comfort.

Keywords: You Only Look Once (YOLO), Canny edge detection, Euclidean distance and Morphological Operations.

Abstract: This project presents the design and implementation of a student feedback system utilizing SQL (Structured Query Language) for database management and PHP (Hypertext Preprocessor) for server-side scripting. The system aims to streamline the process of collecting and analyzing feedback from students regarding various aspects of their educational experience. Key functionalities include user authentication, survey creation, feedback submission, data storage, and reporting. The system employs a relational database schema to efficiently organize and manage feedback data, allowing administrators to generate insightful reports for continuous improvement of academic programs and teaching quality. Through a user-friendly web interface, students can conveniently provide feedback, fostering a collaborative and transparent educational environment. This project demonstrates the practical application of SQL and PHP in developing a robust feedback system to enhance the overall educational experience.

Abstract: VirtualTouch – Replacing keys and clicks with swipes and waves. "VirtualTouch" is a pioneering project that intersects human-computer interaction with cutting-edge technology, redefining computer engagement by seamlessly integrating hand gestures. Leveraging computer vision and machine learning, it interprets and responds to gestures, replacing traditional inputs and bridging the gap between physical actions and digital tasks. Beyond convenience, it enhances accessibility for those with physical limitations and fosters immersive experiences in gaming, design, and education. From hardware setup to interface design, its journey involves overcoming challenges with determination. "VirtualTouch" embodies the fusion of human ingenuity and technological advancement, promising a future where swipes and waves replace clicks and keys, enabling a harmonious coexistence of the physical and digital worlds while preserving human expression. Keywords: Gesture Control, Hand Gestures, Computer Vision, Human, Computer Interaction, Virtual Input, Gesture Recognition, Touchless Interface, Machine Learning, Real- Time Interaction, Accessibility Tech. These concise keywords capture the core aspects of our project for easy reference and search. Cite: Madhura Raut, Vipul Chandrakapure, "Virtual Touch: Replacing clicks and keys with swipes and waves", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13419.

Abstract: Phishing websites represent a significant threat to cyber security as they threaten the confidentiality, integrity and availability of both corporate and consumer data. These malicious sites often serve as an entry point for various cyber attacks. Despite extensive efforts by researchers over the years, effective detection of phishing sites remains a challenge. While some advanced solutions show promise, they often require extensive manual engineering of features and struggle to keep up with emerging phishing tactics. Addressing this challenge requires strategies capable of automatically identifying phishing sites and quickly handling new, previously unseen attacks. One promising approach involves leveraging the wealth of data available on websites hosting these malicious activities. Machine learning is proving to be a powerful tool in this endeavor, offering a more automated and efficient approach compared to traditional methods. In our research, we conducted a comprehensive literature review and proposed a new method for detecting phishing websites. This method involves extracting features from web pages and using machine learning algorithms for classification. Using a data set specifically designed for this purpose, we aim to develop a robust and adaptive system capable of accurately identifying phishing attempts, including zero-day attacks. Through this work, we aim to improve cybersecurity measures by providing a reliable method for identifying phishing attempts, including new and previously unseen attacks. By leveraging the wealth of data available on phishing hosting websites, our approach aims to improve detection accuracy and reduce the risk of data breaches. Ultimately, our goal is to strengthen defenses against phishing attacks and protect sensitive information from unauthorized access.

Keywords: Phishing, Malicious, Cyber ​​Security, Threat, Automation, Security Cite: Mr. V. Ravikanth, Madimi Deekshitha, Palla Gnaneswar, Mallepogu Hari, Anumala Dinesh, "PHISHING ALERT USING MACHINE LEARNING", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13420.

Abstract: The advancement of Internet of Things (IoT) technologies has ushered in a new era for Consumer Electronics (CE), characterized by heightened connectivity and intelligence. This evolution enables enhanced data availability and automated control within CE networks, comprising sensors, actuators, and consumer devices. Cu-BLSTM offers advantages in processing sequential data and capturing long-term dependencies, making it a promising candidate for intrusion detection tasks. However, Cu-BLSTM also presents limitations, including high computational complexity and sensitivity to hyperparameters. To provide a comprehensive analysis, this study compares with Convolutional Neural Networks (CNN), Deep Neural Networks (DNN), and Recurrent Neural Networks (RNN) in the context of intrusion detection for smart CE networks. CNNs excel in extracting spatial features from data, making them suitable for certain types of intrusion patterns. DNNs offer scalability and ease of training, which can be advantageous for large-scale deployment scenarios. RNNs, on the other hand, are well-suited for processing sequential data with temporal dependencies. By understanding the strengths of CNN, DNN, and RNN, this research aims to inform the design and implementation of effective IDS solutions tailored to the unique requirements of smart CE networks. Index terms: Consumer Electronics, Cyber Attacks, Deep Learning, Internet Of Things, Intrusion patterns Cite: Mr. K. R. Harinath M. Tech., (Ph.D.), V. GuruBhargavi, S. Javid Basha, S. Shruthi Keerthana, T. Naveena, "An Innovative Intrusion Detection Systems for smart Electronic Consumers", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13421.

Abstract: Malware detection is a critical aspect of cybersecurity, and the integration of machine learning techniques has shown promising results in enhancing detection capabilities. This paper proposes the integration of the Extreme Gradient Boosting (XGBoost) algorithm for detecting malware. XGBoost, known for its scalability and efficiency, has been successfully applied in various domains, including malware detection. By leveraging XGBoost, this research aims to improve the accuracy and efficiency of malware detection systems. Additionally, the XGBoost algorithm has been compared with other machine learning models like Random Forest and Adaboost, consistently showing superior performance in terms of detection accuracy. The proposed approach in this paper builds on the strengths of XGBoost to enhance the detection of diverse malware variants. By optimizing XGBoost parameters and leveraging ensemble learning techniques, this research contributes to advancing the capabilities of malware detection systems, thereby strengthening cybersecurity measures in the face of evolving cyber threats.

Keywords: Malware Detection, Xgboost, Machine Learning, Dynamic Analysis, DT,KNN,RF, Static analysis.

Abstract: WhatsApp has become integral to modern communication, yet managing unwanted messages and group notifications presents challenges for effective analysis. The model present in paper will address this issue by developing a robust chat analyzer capable of handling such content. Utilizing Python libraries like pandas, seaborn, and matplotlib, alongside advanced natural language processing, the analyzer identifies and filters out irrelevant messages and group notifications. A preprocessing module ensures that subsequent analysis focuses on meaningful conversations, while sentiment analysis provides insights into user interactions. Deployed as a user-friendly application, the analyzer offers comprehensive visualization and statistical analysis of chat data. Through interactive features, users gain valuable insights into conversation dynamics while efficiently managing unwanted content. Incorporating machine learning and sentiment analysis, this project presents a versatile solution for WhatsApp conversation analysis, empowering users to extract meaningful information while mitigating the impact of unwanted content. It is deployed on the Heroku web platform, utilizes a combination of Python libraries such as matplotlib, streamlit, seaborn, re, pandas, and concepts of natural language processing. This amalgamation of machine learning and NLP techniques enables the tool to import WhatsApp chat files, analyze them, and generate various visualizations, enhancing comprehension of the data.

Keywords: WhatsApp, Chat Analysis, Emoji Analysis, Emotion Analysis, Sentiment Analysis, Preprocessing, Natural Language Processing, Data Visualization, Machine Learning, Python Libraries, Pandas, Seaborn, Matplotlib, Text Analysis, Behavioural Analysis, Group Notifications, Unwanted Messages, User Interaction. Cite: Sravanthi D, Lepakshi Reddy S, Vittal Sai C, Jahnavi S, Vamsi C, "ChatProbe Profiling WhatsApp Conversations Using Machine Learning Approaches", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13423.

Abstract: Creative Visionaries uses machine learning and open CV library functions for image processing to convert original image to cartoon for providing fun, and creative images. Beyond this it provides privacy for the image to prevent from morphing using watermark techniques i.e., invisible watermark technique which hides the watermark in the image and prevent from morphing. OpenCV functions are used for edge detection, color quantization and feature extraction, it maintains a balance between preserving image details. This generated cartoon image is used in movies, art, online content and useful for advertisements which attracts the people.

Keywords: Edge Detection, Color Quantization, Bilateral Filtering, Cartoon Image, Watermark Image. Cite: Mr. C. Hrishikesava Reddy*, R. Preethi, D. Sreeya, G. Manoj Kumar, B. Pavan Kumar, "Creative Visionaries Through Machine Learning", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13424.

Abstract: The employee management is crucial for organizational success, and software solutions can greatly support this process. Employees Management Software enables employers to efficiently: - Manage comprehensive employee records (e.g., departments, employee list, salary, leaves) - Add, edit, and remove employee information easily - Establish and modify employee positions - Transfer employees between positions without data re-entry - Identify duplicate positions or employee records for data accuracy - Assign tasks and track employee progress for performance monitoring Cite: Mr. Shreyash Shree Kadam, Mr. Onkar Santosh Jadhav, Prof. Rahul Patil, "EMPLOYEE MANAGEMENT SYSTEM", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13425.

Abstract: "Data Driven Roads: A Connected and Secure Vehicle Mobility Network" is designed to transform the navigation of emergency vehicles in urban congestion, addressing a global challenge that affects timely medical and emergency interventions in cities. Traditional reliance on sirens and manual traffic control often proves inadequate in high-density areas, leading to this project's innovative, data-driven approach aimed at enhancing the efficiency of emergency services. By leveraging sophisticated mobile application technology integrated with the urban infrastructure, the project aims to close the gap between the critical need for quick emergency responses and the existing bottlenecks. The core goal is to design, develop, and implement a state-of-the-art mobile application that supports instantaneous data sharing, facilitating optimized routing for emergency vehicles in real time. This endeavor is built on a foundation of software development expertise, complemented by detailed simulation testing in urban-like environments. A significant emphasis is placed on developing a user-friendly interface that smoothly aligns with current traffic management frameworks, bolstered by reliable real-time communication protocols to ensure the swift and efficient passage of emergency vehicles through city landscapes. "Data Driven Roads" is committed to drastically reducing the time it takes for emergency responses, thereby elevating public safety and enhancing urban mobility. This initiative not only tackles a critical social challenge but also proposes a flexible model that can be tailored to various urban settings. With its trailblazing application of technology to facilitate critical emergency services, the project positions itself as a leader in urban mobility innovations, representing a crucial advancement towards the realization of smarter, safer urban environments.

Keywords: Urban congestion, emergency vehicles, data-driven solution, mobile application technology, urban infrastructure integration, real-time data exchange, dynamic route optimization, simulation testing, user interface, traffic management systems, real-time communication protocols, public safety, urban mobility, scalable model, smart cities.. Cite: Arpitha M, Shashank C, Shravan Vaidya, Sudasrhan, "Data Driven Roads : A Connected And Secure Vehicle Mobility Network", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13427.

Abstract: In the pursuit of accurately predicting stock market movements, researchers have increasingly turned to advanced machine learning techniques. This study explores the application of Long Short-Term Memory (LSTM) networks to stock price forecasting, leveraging financial data obtained through the  y-finance library. The research methodology involved ingesting historical stock price data, macroeconomic indicators, and other relevant features into an LSTM network architecture. The model was trained to learn the complex temporal dependencies and patterns inherent in the financial time series data, with the goal of generating accurate buy/sell signal predictions. Experimental results on a diverse portfolio of publicly traded stocks demonstrated the superior performance of the LSTM-based approach compared to traditional time series analysis methods. The model was able to capture subtle market dynamics and achieve notably higher accuracy in forecasting future stock price movements. The findings of this study suggest that the integration of LSTM networks and accessible financial data sources, such as y-finance, can provide a powerful tool for investors and traders seeking to optimize their investment strategies. The technique holds promise for further advancements in the field of automated financial decision-making. The technique holds promise for further advancements in the field of automated financial decision-making. This final sentence suggests that the LSTM based approach has the potential to drive further progress in the area of automated financial analysis and decision-making. By utilizing readily available tools and technologies, one can embark on their journey towards stock market prediction, potentially making informed investment decisions based on  y-finance sentiment analysis and long short-term memory modelling. This research provides a valuable tool for market participants to gain a deeper understanding of market sentiment and make data-driven investment decisions.

Keywords: Stock Market, Machine Learning, Analysing, Prediction & Education etc. Cite: Prof. Raksha Kardak., Aniket Chandore, Piyush Borkar, Pranay Alikane, Priyanshu Ramteke, Purushottam Kakde, "STOCK TRADE PREDICTION USING Y-FINANCE AND LONG SHORT-TERM MEMORY (LSTM).", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13428.

Abstract: Software development is The primary purpose of making the online property management system is to create an automatic online based system which will provide an easy and an alternate way to book a property. It which provides a friendly format for purchasing and selling properties. User provided with an account. Users can search and browse for property during this application. Users allow to book the property online. This application mainly concentrates on maintaining and managing the small print of the property. This application deals with buying and selling the homes, lands, commercial properties. It provides functionality for the seller, authorize them to log into the system and add new advertisements or delete existing ones.

Keywords: Property Management, efficiency, confirm booking.

Abstract: Through the integration of multiple modalities, fostering a more immersive and effective educational experience. In the context of this study, "multimodal" refers to the incorporation of various media elements such as text, images, videos, interactive simulations, and social interaction tools within the online learning environment. We explore how the judicious use of these modalities can address the prevalent issues of disengagement and cognitive overload often associated with online learning. The research methodology includes the development of a customized online learning platform that strategically integrates these multimodal elements. To evaluate the impact of this approach on engagement, retention, and overall learning outcomes, we conducted a series of controlled experiments with learners from diverse backgrounds. The findings reveal significant improvements in learner engagement, information retention, and satisfaction levels when compared to traditional online learning methods. Furthermore, this research sheds light on the importance of adaptability in multimodal content delivery, tailoring the learning experience to individual preferences and learning styles. This research paper not only contributes to the ongoing discourse on online education but also provides practical insights and guidelines for educators, instructional designers, and e-learning developers seeking to create more engaging and effective online learning environments. In the realm of digital education, this multimodal approach shows potential for enhancing education quality and accessibility, paving the way for a brighter future for online learners worldwide.

Keywords: Onlinelearning, engagement, multimodal approach, learning outcomes, customized online learning platform, adaptability.

Abstract: This research papers gives an analysation over the features of heart disease analysis module that deals with a wide range of complications. Heart can be called as the most crucial and useful organ in the body of any organism especially in human body. It plays a very important and purpose full role. Diagnosing the heart is very important and regular check-ups regarding heart and it’s related diseases can be helpful in living a healthy life. To tackle such heart complications and to help people be cautious regarding their heart conditions there exists an essential urge of heart disease prediction system. Machine learning provides the noticeable support in finding any type of the event which demands a structural training from the naturally occurring events. For the purpose of disease analysis, machine learning is thought to be the most well-known and notorious platform, while the convenience of utilising it along with computational domination in the array of modules that grabs the attention millions of patients all over globe, which also demands the several technical concerns in different platforms. This research article aims to analyze the predictive power of several machine learning algorithms for cardiac disorders.

Keywords: Heart Disease.prediction, machine learning, analysis

Abstract: Electric vehicles (EVs) are gaining traction globally as a sustainable solution to transportation needs, promising reduced reliance on fossil fuels and lower emissions. However, the challenge of convenient and eco-friendly charging infrastructure persists. This project addresses this challenge by introducing a dynamic electric vehicle charging system powered by solar energy. The system leverages a 12V solar panel to harness renewable energy, converting it into electrical power for charging EV batteries. What sets this system apart is its wireless transmission technology, enabling continuous charging while the vehicle is in motion. This eliminates the need for external power sources or stopping for charging, enhancing the convenience and efficiency of EV usage. Central to the system's operation is the Arduino UNO microcontroller unit, which manages the charging process and ensures optimal energy transfer. Real-time data on charging status and performance are displayed on a 16 X 2 LCD display, providing users with valuable insights. Key components such as the DC converter, transmission circuit, and copper coils are integrated seamlessly to facilitate efficient charging. This integration not only ensures a smooth charging experience but also underscores the system's sustainability and environmental friendliness. In summary, this project presents a holistic solution to the challenges of EV charging infrastructure. By harnessing solar energy and employing wireless transmission technology, it offers a sustainable, efficient, and convenient charging option for electric vehicle users, ultimately contributing to a greener and more sustainable future.

Keywords: Solar Energy, Wireless Charging, Electric Vehicle, Arduino UNO, DC Converter, Transmission Circuit, Sustainable Charging, Dynamic Charging System

Abstract: Skin conditions are very infectious and harmful. One of the several disorders brought on by bacteria, viruses, fungi, or allergies, dermatosis is the most prevalent. could result in. Skin illnesses are known as dermatoses. Certain skin conditions can only be treated by a dermatologist; other skin conditions cannot be treated with the right drugs. Rashes, inflammation, itching, and other skin abnormalities can also bebrought on by skin illnesses. The goal of this project is to use image processing to identify skin conditions. In this procedure, we take a picture of the affected area, analyze it using image processing techniques to determine what kind of sickness it is, and then apply CNN for real-time analysis.

Keywords: CNN, Image Processing, Deep learning, Machine learning

Abstract: Underground mining operations pose significant safety risks, making early detection of hazards and real-time monitoring of worker health crucial. This abstract presents a novel Undermining Detection Robot (UDR) designed to enhance safety in underground mining environments. The UDR employs Arduino microcontrollers to interface with a suite of sensors, including metal sensors, fire sensors, gas sensors, ultrasonic sensors, and moisture sensors. These sensors provide real-time data on potential hazards such as gas leaks, fires, and unstable underground conditions. Additionally, the robot incorporates a water pump to address moisture related issues that may arise in mining operations. The UDR is equipped with an ESP32 CAM module, stream video from the mining site. This feature enhances remote monitoring and situational awareness for mine operators. The data collected from these sensors and the ESP32 CAM are transmitted to the cloud-based IoT platform, ThingSpeak, for real-time data analysis and visualization. The UDR also integrates sensors for monitoring personnel health parameters, including heartbeat, temperature, and Spo2 levels. This functionality ensures that workers' well-being is constantly monitored, and any anomalies or emergencies are promptly detected. The final layer of innovation in this system is the application of machine learning using Python. The collected data is analyzed using machine learning algorithms to predict potential safety hazards or health related issues. These predictions are then used to trigger immediate responses or alert personnel to take appropriate actions.

Keywords: UDR, sensors, ESP32CAM , Thingspeak ,ESP32,Arduino meg

Abstract: A last mile connectivity app provides efficient and effective connectivity for the final leg of the transportation network for the users. This kind of app studies the route an individual has to travel and then determines his/her path in the most convenient way possible. Keeping the users updated about the modes of transport in his/her way is the main aim of this app. EasyTrans plays a crucial role in determining the overall quality and accessibility of services in a networked society. The use of this application have a high perceived time savings among users, and it is found that there is a measurable modal shift from personal vehicles to these solutions for bridging the last mile gap.

Keywords: last mile connectivity, rapid transit systems, accessibility, intermediate public transport, feeder services

Abstract: We offer a novel Convolutional Neural Network (CNN) method designed exclusively for detecting diabetic retinopathy (DR) in visual pictures. The existence of DR at an early stage has an impact on the effectiveness of treatment. Ophthalmologists commonly physically identify DR in retinal pictures; our objective is to do it correctly. To improve accuracy and avoid overfitting, we develop a specialised CNN architecture, clean the data, and employ a customised dataset. We increase actuation capabilities and hyperparameters through meticulous design. Our technique beats earlier frameworks in terms of accuracy, review, F1-score, and ROC-AUC, according to our research. By seeing the CNN highlight maps, one may have a better grasp of diagnosis. Our findings show that updated deep learning models might be used in restorative imaging to deliver rapid and accurate DR diagnosis, save medical staff workload, and possibly even safeguard patients' eyesight.

Keywords: Diabetic Retinopathy, CNN Architecture, Image Classification, Deep Learning, Customized Model, Medical Imaging

Abstract: This research project focuses on the development of a comprehensive data processing framework tailored for advanced machine learning applications in the domains of classification, regression, and clustering. The primary objective of this endeavor is to empower users with the tools and methodologies necessary to effectively explore and exploit their datasets for predictive and exploratory data analysis. The project leverages state-of-the-art machine learning algorithms and data processing techniques to facilitate the training, testing, and evaluation of machine learning models across diverse application scenarios. By offering a flexible and user-friendly environment, it enables researchers and data practitioners to harness the full potential of their data. Key components of this project include data preprocessing, feature engineering, and model evaluation. Data preprocessing encompasses various techniques such as data cleaning, transformation, and normalization to ensure data quality and consistency. Feature engineering involves the creation of meaningful and informative features that enhance the performance of machine learning models. Model evaluation incorporates a variety of metrics and visualization tools to assess the effectiveness and robustness of the trained models.

Keywords: DatavisualizationLiterary,narrativedata,Database,managementsystem,Neo4J,Graphdatabase,Webapplication,Streamlit,Pyvis,Relationshipsbetweenentities,Bidirectional Encoder (BERT)

Abstract: One of the fundamental five senses that are essential to daily life is hearing. Despite sufficient mindfulness, society has a disgrace around hearing misfortune. It is one of the critical issues in this present reality and is expanding dramatically. Early recognition and intercession are the method for forestalling and treating this issue. Access to hearing health care is becoming an increasingly pressing issue on a global scale, as it is estimated that by 2050, around 900 million individuals will be affected by disabling hearing loss. The obstacles that pure-tone audiometry encounters encompass large and expensive equipment, restricted accessibility, scarcity of professionals, and ineffective data management. These challenges impede the widespread acceptance and holistic patient care. App-based solutions revolutionise pure-tone audiometry by enabling early detection of hearing impairments through accessible and convenient testing. They offer users the flexibility to undergo tests regularly and remotely, facilitating swift interventions when necessary. Beyond convenience, these solutions provide robust data tracking and analysis capabilities, enhancing our understanding of hearing health and enabling personalized care strategies. Integrating app-based technologies represents a pivotal advancement, addressing key challenges in the field such as accessibility and data management. Ultimately, these innovations contribute to improved hearing health outcomes on a global scale.

Keywords: Pure Tone Audiometry, Decibel, Hearing loss, Frequency

Abstract: The advent of blockchain technology has sparked interest in its potential application to various domains, including voting systems. This paper proposes a novel online voting system leveraging blockchain technology to address critical challenges such as security, transparency, and verifiability in electoral processes. Our system employs a distributed ledger to record votes securely, ensuring immutability and tamper resistance. Smart contracts are utilized to automate voting processes, enhancing efficiency and reducing human errors. We present a comprehensive architecture for our blockchain-based online voting system, encompassing voter registration, ballot creation, voting, and result tabulation phases. A user-friendly interface ensures accessibility for voters, while robust authentication mechanisms prevent unauthorized access and ensure the integrity of the electoral process

Keywords: Blockchain, Consensus, E-voting, Security, Transparency, Smart-contract, Ethereum.

Abstract: In recent years, significant progress has been made   towards human-like machine comprehension. However, it's crucial to be aware that teaching a computer to think like a person is a very challenging undertaking. Automation and how we interact with computers are becoming more fun as computer vision technology advances. The Viola Jones approach, which makes it easier to recognise persons in photos, is discussed in this research. No matter the illumination, we have the computer set up to perform this task automatically. In experiments, they studied the Viola-Jones Cascade Object Detector. In order to identify particular facial features, this detector employs a variety of filters and traits.

Keywords: SURF, Feature Detection, Face Recognition, Face Detection, Viola-Jones algorithm.

Abstract: The research paper explores the realm of AI-driven insights and data visualization, focusing on the utilization of LIDA (Language-Integrated Data Analysis) as a powerful tool for facilitating data analytics. In today's data-centric world, the ability to extract meaningful insights and communicate them effectively is paramount for informed decision- making. Our project aims to democratize the field of data analysis by providing an intuitive and inclusive platform accessible to users of all technical backgrounds. LIDA leverages cutting-edge Natural Language Processing (NLP) and Machine Learning (ML) techniques to enable users to effortlessly upload CSV files and engage in natural language conversations to extract insights, generate visualizations, and receive predictive analytics. The methodology encompasses comprehensive data collection and preprocessing, deploying robust NLP models for language comprehension, and integrating ML algorithms for data analysis. The chatbot's interface prioritizes user-friendliness, offering an intuitive environment for data upload, user interactions, and actionable insights. Through real-world case studies and examples, we demonstrate the effectiveness of LIDA in generating actionable insights and facilitating data-driven decision-making. The research contributes to bridging the gap between data expertise and non-technical users, empowering a broader user base to harness the potential of artificial intelligence in data analytics.

Keywords: NLP, LIDA, Machine Learning, Insights, CSV File, Data Analytics, Data Processing, Data Visualization

Abstract: This research paper introduces LIFE SYNC, an innovative AI chat bot that improves user interactions through speech recognition, robotics, natural language processing, and real-time editing. This project prioritizes a thorough requirements analysis covering user, functional, and non-functional aspects with the aiming to develop a sophisticated voice-activated assistant. Using cutting-edge technologies including natural language processing, robotics, and a robust chat bot framework, LIFE SYNC enables efficient data storage and organization. LIFE SYNC's architecture integrates voice recognition, data management, conversational interaction, and user interface modules to ensure smooth operation. This system places emphasis on security and privacy and has implemented measures to safely handle user data. This implementation includes various components such as the main file, dashboard, chat screen, home page, profile screen, Open AI service, and features such as API key integration, audio processing, font management, and the use of external libraries. all controlled by the Android manifest file. LIFE SYNC offers benefits such as improved user accessibility, real-time editing, and efficient data management, making it suitable for a range of uses, including customer support, virtual personal assistance, e-commerce, product recommendations, and healthcare support. Future prospects are outlined in the study paper, and they include chances for collaboration with third-party platforms, multilingual support, connection with new technologies, and interaction with smart homes. LIFE SYNC represents a convergence of advanced technologies must meet the requirements of diverse users in an evolving digital environment.

Keywords: Speech Recognition, Natural Language Processing, Machine Learning, Efficient Data Management, User Interface Integration, Virtual Personal Assistant.

Abstract: Bloom's Taxonomy is a framework used by educators to categorize the learning objectives that they assign to their students. This taxonomy's cognitive domain is intended to confirm a student's cognitive proficiency in a written test. Teachers may occasionally find it difficult to determine whether the exam questions they create adhere to Bloom's taxonomy requirements at various cognitive levels. Based on this taxonomy, this research suggests an automated examination question analysis to identify the relevant category. This rule-based method uses Natural Language Processing (NLP) approaches to find significant verbs and keywords that could help determine a question's category. The topic area of computer programming is the main emphasis of this work. Currently, the research uses a set of 100 questions, comprising 30 test questions and 70 training questions. According to preliminary findings, the guidelines could be able to help candidates accurately identify the Bloom's taxonomy category in test questions. By utilizing the hierarchical structure of Bloom's cognitive domain, automatic question paper production using Bloom's taxonomy can generate questions with different levels of complexity and cognitive ability. After analyzing the learning objectives or content using algorithms and natural language processing, the system creates pertinent questions that correspond with the levels of Bloom's taxonomy, which include knowledge, comprehension, application, analysis, synthesis, and assessment. This method aids teachers in developing thorough, well- balanced exams for pupils that encourage critical thinking and deeper comprehension.

Keywords: Bloom’s Taxonomy, Natural Language Processing(NLP)

Abstract: Skin cancer ranks among the most prevalent forms of cancer globally, contributing to millions of fatalities. Its primary cause lies in the uncontrolled mutation growth within DNA. Early detection is pivotal for enhancing treatment success rates. Modern medical practices leverage technology extensively, with intelligent systems aiding in the analysis and classification of skin conditions. Detecting infection rates accurately poses challenges due to the intricate texture of skin and the visual similarities of diseases. This study aims to analyze biomedical datasets containing pre-existing illness data to develop an effective method for distinguishing skin cancer as either malignant or benign. Leveraging ResNet-50 deep learning architectures, we classify the dataset. With a dataset comprising training images, our model achieves an accuracy of 86.66%. This accuracy may further improve with increased epochs.

Keywords: Skin Cancer, Intelligent-systems, Resnet-50, Malignant, Benign.

Abstract: Regular online card exchanges have expanded as a result of innovative headways in ranges like e-commerce and monetary innovation (FinTech) applications. Credit card extortion has expanded as a result, having an affect on card backers, retailers, and as well as banks. In this manner, making frameworks to ensure the astuteness and security of credit card exchanges is pivotal. In this ponder, we utilize skewed real-world datasets from European credit cardholders to build a machine learning (ML) based system for credit card extortion discovery. We re-sampled the dataset utilizing the Synthetic Minority over- sampling method (SMOTE) in arrange to address the issue of lesson lopsidedness. We evaluated this system with the taking after machine learning methods: Extreme Gradient Boosting (XGBoost).

Keywords: SMOTE, credit card, data resampling, fraud detection, XGBoost, machine learning.

Abstract: With the expanding utilize of advanced pictures in different applications, the issue of picture fraud has ended up more predominant than ever One of the greatest issues these days is picture frauds or control using different procedures In this paper, we propose a novel advanced picture fraud location framework based on Convolutional Neural Systems (CNNs) that can identify different sorts of picture controls, counting copy-move, grafting, and correcting. Our proposed framework coordinating Mistake Level Investigation (ELA) with profound learning strategies to supply a more exact and dependable arrangement to the issue of computerized picture imitation location. We assessed the proposed framework on a dataset of real-world pictures and accomplished a tall location precision of 93% Our framework outflanked existing strategies for picture imitation location and illustrated its potential for different applications, counting forensics, security, and computerized picture investigation. A convolutional neural organize that has been appeared successful for picture preparing is utilized at first. Generally, the proposed CNN-based picture imitation location framework offers a vigorous and successful arrangement to the developing issue of picture control and fraud in today's visual media scene. The execution of the proposed strategy is tried quantitatively, and picture alteration is distinguished

Keywords: Digital image analysis, Convolutional neural network (CNN), Error Level Analysis (ELA), Image processing.

Abstract: Now a days, road accidents and traffic issues are on the rise. Video-based vehicle detection technology can gather valuable data from video frames, like vehicle speed, type and license plates cost effectively and efficiently. This data can improve traffic management and enhance safety. Traditional speed detection systems still rely on traditional algorithms like YOLOv5 and SSD that lack in accuracy. Our proposed system uses cutting-edge technologies in machine learning and video steaming analysis. The system integrates YOLOv8 for precise vehicle detection, DeepSORT for robust tracking of vehicles, and EasyOCR in conjunction with YOLOv8 for accurate license plate number detection. On detecting the vehicle the speed of each vehicle is estimated and it’s details are noted upon speed violation. Integration of these technologies improves traffic monitoring and reduces road accidents. The data collecting by these can later be used for alerting the drivers and as well as control teams.

Keywords: Vehicle detection, tracking, Vehicle speed detection, video streaming analysis, DeepSORT, YOLOv8 EasyOCR.

Abstract: The Information Technology (IT) professionals often face high-stress levels due to the demanding nature of their work. Chronic stress results in cancer, cardiovascular disease, depression, and diabetes. Developing robust methods for the rapid and accurate detection of human stress is of paramount importance. This research aims at developing a stress detection and management system specifically tailored for IT employees. Machine Learning algorithms and Deep Learning techniques like Convolutional Neural Networks are used for face emotion detection. Image Processing is used at the initial stage for detection. The existing methods for real time face emotion recognition and worker stress analysis has a draw-back that there is no live detection. Additionally, the existing system lacks a contextual data component to account for external factors influencing stress. The proposed system includes the real time live cam detection of Face emotion recognition and Worker Stress analysis and periodic analysis of employees and detecting mental stress levels through seamless integration of deep learning models for emotion detection. It provides a comprehensive understanding of employee well-being. The ultimate goal of our research is to identify emotion levels in employees, providing a foundation for stress management and thus enhance employee well-being and ultimately improving individual overall quality of life by enabling early stress detection.

Keywords: Image Processing; Convolutional Neural Networks; Stress

Abstract: In todays realm cryptography plays a role, across various sectors. Image encryption stands out as an aspect in safeguarding data due to its ranging applications in areas like defense, multimedia, healthcare and more. This article introduces an image encryption technique designed for both grayscale and color images using the Tangential Delay Ellipse Reflecting Curve System (TD ERCS) chaotic map system and DNA coding. The chaotic map is employed to shuffle positions for confusion and mask image creation while DNA coding alters pixel values for diffusion. Through evaluation the proposed method demonstrated high mean square error and low peak signal to noise ratio, near zero correlation, a high rate of pixel changes consistent average intensity value changes, as well as resilience, against noise and data loss attacks. Furthermore decryption can be carried out without compromising the image quality. Keywords: Image encryption, Color image, Chaotic map, DNA coding, AES, IDEA, DES

Abstract: Malware detection is a critical component of modern cybersecurity, as malicious software poses a substantial threat to the security and privacy of individuals and organizations. Traditional signature-based approaches to malware detection have limitations in identifying new, previously unseen malware variants. Machine learning has emerged as a powerful tool in this domain, offering the ability to detect malware based on patterns and behaviour’s rather than relying solely on known signatures. These abstract highlights the key aspects of using machine learning for malware detection. Machine learning algorithms are capable of analysing large datasets of file characteristics, network traffic, and system behaviours to identify subtle and evolving patterns associated with malware. By employing techniques such as deep learning, decision trees, and support vector machines, these algorithms can generalize from labelled training data to make predictions about the nature of unknown files or activities. Additionally, feature engineering and feature selection processes enhance the ability of machine learning models to distinguish between benign and malicious entities effectively. The dynamic nature of malware necessitates real-time or near-real-time detection methods. Machine learning enables the development of predictive models that continuously adapt to new threats, making it possible to stay ahead of evolving malware variants. Moreover, the integration of machine learning with other security measures, such as anomaly detection and threat intelligence, further enhances the overall efficacy of cybersecurity systems.

Keywords: Malware detection, Machine learning, Behavioural analysis, Decision trees, Feature engineering.

Abstract: Mobile internet will help the farmers to sell their products directly to consumers and food processing industries. This paper provides market information to a farmer using its easy interface on the mo- bile application. The mobile application is intended to be used for fast and updated information delivering system for farmers. Also, it has native language support to make the transaction easy for farm- ers. The mobile application treats farmers as a seller and a buyer. The intention behind this paper is to help farmers so they buy or sell their agriculture goods and products. Market prices provided by data.gov.in lets the system to keep the selling and buying prices in control. As the products are to be browsed and there may be plenty of products for the user. To make browsing easy many filters can provide. Farmers face many problems while selling their goods and products, this system promises to provide an easy and recreational way to sell the products. The system lets the farmers to sell goods at a reasonable price and makes business even fair and transparent. Consumers are the opposite side of the same coin. This system lets consumer to choose from a wide variety of products, select the prod- uct as per their requirement and also to apply price filters. Location is a one of parameter for consumer and producer while selling or buying their product it will helps the user to get the product nearby their location. The basic objective of the system is to considers every one need and full fills their requirement with fair and transparent agriculture business.

Keywords: Agriculture, Farm Assistant, Dealer, Customer, Shopping cart, Java, Android, Agricultural Products

Abstract: Numerous industries, including autonomous vehicles, security, and image analysis, use object detection, a crucial computer vision problem. You may use TensorFlow, one of the top deep learning frameworks, to do object identification, and this detailed guide will walk you through every step of the procedure. We'll walk you through the initial setup before demonstrating how to choose pre-trained models from the TensorFlow Model Zoo and help you customise them for your particular object identification task. You'll gain knowledge of dataset preparation, data augmentation, and crucial model-training procedures. The complexity of model evaluation is also covered in this book, which will assist you in assessing the effectiveness of your object detection model using metrics like Mean Average Precision (mAP). We offer insight into typical problems and the best solutions, with a practical focus. Finally, we look at the procedure for reasoning about fresh picture or video streams using your learned model. From setup to inference, the entire object detection process will be thoroughly explained to you, giving you the information and abilities you need to use TensorFlow to meet your object detection needs.

Keywords: Object Detection, TensorFlow, Setup, Inference, Deep Learning, Comprehensive Guide

Abstract: The agriculture sector, a linchpin of global economies, finds itself at a crucial juncture where entrenched traditions intersect with the transformative waves of technology. This research endeavorsto address pressing challenges within agriculture, chief among them being low yields and the distressing plight of farmers. The proposed solution takes the form of a neural network-based agricultural yield forecast system, wielding the power of modern technology to offer innovative pathways forward.Harnessing the potential of a user-friendly smartphone application, the system establishes direct communication channels with farmers. Leveraging GPS technology, it not only identifies their precise locations but also captures crucial information vital for precision agriculture. Machine learning techniques, prominently featuring the Random Forest algorithm, form the backbone of this system. Impressively, it achieves an accuracy rate of 95% in predicting agricultural yields, marking a significant leap forward in predictive analytics for farming outcomes.This novel approach seeks to narrow the chasm between conventional farming methods and the rapid strides of contemporary technology.

Keywords: Crop Recommendation, Machine Learning, Random Forest, Decision Tree, Logistic Regression, XGBoost, Data Analysis, Data Visualization

Abstract: Animal Vehicle Collision, commonly called as roadkill, is an emerging threat to humans and wild animals with increasing fatalities every year. Amid Vehicular crashes, animal actions (i.e. deer) are unpredictable and erratic on roadways. This paper unveils a newer dimension for wild animals’ auto- detection during active nocturnal hours using thermal image processing over camera car mount in the vehicle. To implement effective hot spot and moving object detection, obtained radiometric images are transformed and processed by an intelligent system. As human populations expand and encroach upon natural habitats, conflicts between humans and wildlife become increasingly common. To mitigate the risks associated with wild animal intrusions into human settlements, an intelligent and proactive intrusion detection system is essential. This study proposes a novel approach to wild animal intrusion detection using deep learning techniques. The proposed  system leverages Convolutional Neural Networks (CNNs) to analyze images captured by surveillance cameras placed in strategic locations. The deep learning model is trained on a diverse dataset of wildlife images to enable accurate identification and classification of different species.

Keywords: Wild Animal Intrusion Detection System, Smart Protecting, Animal Detection , Precision Farms from Animals , Protection of Animals.

Abstract: Lung cancer is one of the leading causes of cancer-related deaths worldwide. Early detection of lung cancer is crucial for successful treatment and improved survival rates. In this paper, we propose a system for lung cancer detection using digital image processing and machine learning techniques. The proposed system uses digital image processing techniques to segment lung nodules from computed tomography (CT) scans. Features are then extracted from the segmented nodules using texture analysis and shape analysis. These features are used to train a machine learning classifier that can differentiate between malignant and benign nodules. We evaluated the performance of the proposed system using a dataset of 300 CT scans from the Lung Image Database Consortium (LIDC). Our results show that the proposed system achieved an accuracy of 91.67% in detecting lung nodules, outperforming other state-of-the-art approaches. Overall, the proposed system has the potential to improve the accuracy and efficiency of lung cancer detection, leading to earlier diagnosis and better treatment outcomes.

Keywords: Feature Extraction, Adaptive Thresholding, Matching, Multi-Label Classification, CT (computed Tomography), Image Processing, Machine Learning, Convolutional Neural Network (CNN), etc

Abstract: With the snowballing intricacy and erudition of cyber intimidations, traditional security measures have become insufficient to combat modern cybersecurity challenges. As a result, the amalgamation of artificial intelligence (AI) into security systems has arisen as a promising tactic to boost threat detection and mitigation. This article investigates the advancements in AI-based security and threat detection, examining various AI techniques, their applications in cybersecurity, challenges, and future directions. By leveraging AI, organizations can improve their ability to successfully identify, evaluate, and counteract cyberthreats, enhancing overall security posture

Keywords: AI, cybersecurity, threat detection, machine learning, deep learning, neural networks, anomaly detection, security systems.

Abstract: This initiative intends to streamline cell financial transactions by using blockchain technology for more appropriate security. Users can safely manage finances, communicate, and access financial offers on their mobile gadgets. The integrated blockchain provides secure and transparent transactions, with a focus on strong authentication mechanisms to protect sensitive records. The challenge demonstrates the efficiency and effectiveness of blockchain in cell financial transactions for high security and record protection.

Keywords: Energy efficient algorithm; Manets; total transmission energy; maximum number of hops; network lifetime

Abstract: To keep an eye out for criminal actions like theft, data modification, and system disruption on one or more computers, we develop an intrusion detection framework. Digital attacks that are dynamic and complex are difficult for traditional intrusion detection systems to detect. But utilizing reliable methods, such various kinds of artificial intelligence, can raise detection rates, lower false alarm rates, and provide affordable solutions. Particularly in data mining, continuous pattern analysis, categorization, aggregation, and real-time data processing are made possible. This research study offers a targeted analysis of the literature on enhanced intrusion detection techniques using data mining and artificial intelligence.[2] In order to provide an analysis, synthesis, and succinct overview of their contents, we identify pertinent publications based on the volume of citations or emerging trends. We also emphasize data's crucial significance in data mining and artificial intelligence.[4]

Keywords: Intrusion detection framework, Artificial Intelligence, data mining, cyber security, cyber resilience

Abstract: While India’s population continues to increase in today’s world, the health of our country is also under threat. Progress will increase in this word ,but at the same time ,the health of our country is at risk. So subject is to provide clean toilets. All public toilets must be clean and hygienic. In our system ,we decided to take care of the cleaning of a toilet. It can avoid many diseases. It can lead to the management of toilets. Therefore, our evolution id to use safe and clean toilets. Using the Arduino concepts which uses different sensors like MQ-135 Sensor,MQ-8 Sensor, RFID Reader, RFID Tag ,MQ 4,Arduino,DHT-11 Temperature and Humidity Sensor. .By using these sensors, we can create the smart toilets. Proposed system cleans the public toilet with the help of Arduino technology. TheArduino-based E-Toilet system, mainly deals with solving the problem of the unhygienic condition of public toilets. The hardware kit has attached in the toilet with location ,Ki-Id and Cleaning Boy details. To maintain the periodicity of cleanliness level different kind of sensors are used. A database is maintained which gives all the notifications to authorities of cleaning department of municipal corporation on a web page. MNC views cleaning logs and Uncleaned toilets. System also provide RFID reader. When the sensor value crossed threshold values then smell sensor detect unclean toilet If toilet is unclean then cleaning boy read there RFID tag . Hardware kit has RFID tag that contains a unique ID scanned by the cleaning boy. Kit data save on server. When the RFID Tag is detected by the Cleaning boy, the system will get all sensors value. Cleaning boy Clean Toilet and After Toilet cleaning read RFID Tag to get all sensor values to view toilet conditions. The advantages of the system are that it reduces the labour work and its working is flexible. In India this type of project is not implemented. It is affordable for municipal corporation for its implementation. Keywords- MQ-135 Sensor, MQ-135 Sensor, Rfid Reader, Rfid Tag, MQ-4, Arduino, DHT-11 Temperature Humidity Sensor, MNC.

Abstract: Road traffic safety is a crucial concern around the world, the probability of accidents and collisions rises as the number of cars on the road grows. Object detection technology has developed as an effective technique for enhancing road traffic safety. This research provides a detailed evaluation of the most current breakthroughs in object recognition systems and their applications in road traffic safety. The paper opens by providing an overview of the issues and risks involved with road traffic, highlighting the importance of enhanced safety measures. It then digs into a full review of object identification strategies, ranging from traditional methods to cutting-edge deep learning models, demonstrating their capacities to identify vehicles, pedestrians, cyclists, and other road items.It investigates how these technologies improve real-time monitoring, collision avoidance, and traffic management. Furthermore, the article looks into object detection for traffic law enforcement and monitoring, emphasizing its significance in improving security and lowering accidents. It outlines prospective future research directions, such as the development of powerful, real-time object detection systems and their application to smart city initiatives.

Keywords: Real-time object detection, road traffic safety, bounding boxes,intersection over Union (IOU), Anchor boxes, non-max Suppression.

Abstract: Maize stands as a vital agricultural crop worldwide, serving as a crucial source of sustenance for humans, livestock feed, biofuel, and a raw material for various products. The detection and management of natural diseases pose a significant challenge for food crops. Swift identification of plant diseases remains a time-consuming and arduous task, particularly for small-scale farmers. Conventional methods and tools lack efficacy, demanding extensive manual labor and time investment. Timely disease detection is imperative for effective treatment and timely implementation of pesticide measures to curb the spread. This research introduces an efficient image classification model based on advanced deep learning techniques, specifically tailored for accurately identifying three prevalent maize leaf diseases. The proposed model employs the Xception model, leveraging transfer learning through pre-trained Xception models for robust feature extraction. The amalgamation of deep features creates a sophisticated feature set, enhancing the model's ability to derive valuable insights from the dataset. With reduced computational costs and the capability to capture essential characteristics, this depth-wise separable Convolutional Neural Network (CNN) exhibits superior efficiency. Comparative analysis against other CNNs, such as EfficientNetB0 and DenseNet121, highlights the exceptional performance of the suggested model, achieving an impressive accuracy of 99.40%. The study underscores the suggested model's superior accuracy and its proficiency in diagnosing various corn leaf diseases. Keywords – Deep Learning, Corn Leaf Diseases, Image Classification, Convolutional Neural Network (CNN), Xception Model, Disease Detection, Agricultural Automation, Precision Farming, Crop Disease Management, Plant Pathology.

Abstract: Every day, a significant volume of textual content is shared online. Sorting through such vast amounts of textual material to find the relevant and irrelevant information is challenging. One area of natural language processing that enables the examination of textual data is sentiment analysis. Since it examines the words and presents the public’s overall viewpoint, it is regarded as an opinion mining technique [1].Sentiment analysis is a domain with three sub-branches: aspect-based, sentence-based, and document-based [2]. Sentences are used to find opinions in sentence-based sentiment analysis. Sentiment analysis of complicated texts is a challenging task. When conducting sentiment analysis on documents, the entire textual Social media provides a forum for public sharing.their opinions and concepts. The most widely used social media sites are Facebook, Twitter, and YouTube, where usersrespond by leaving comments and like the page. Sentiment analysis is a widely utilised tool for analysis these days.

Keywords: toxic comment,Machine Learning,SVM,NLP.

Abstract: Melanoma skin cancer detection at an early stage is crucial for an efficient treatment. Recently, it is well known that, the most dangerous form of skin cancer among the other types of skin cancer is melanoma because it’s much more likely to spread to other parts of the body if not diagnosed and treated early. The non-invasive medical computer vision or medical image processing plays increasingly significant role in clinical diagnosis of different diseases. Such techniques provide an automatic image analysis tool for an accurate and fast evaluation of the lesion. The steps involved in this study are collecting dermoscopy image database, preprocessing, segmentation using thresholding, statistical feature extraction using Gray Level Co-occurrence Matrix (GLCM), Asymmetry, Border, Color, Diameter, (ABCD) etc., feature selection using Principal component analysis (PCA), calculating total Dermoscopy Score and then classification using Convocation neural network (CNN). results show that the achieved classification accuracy is 92.1.

Keywords: Melanoma Skin Cancer, Image Pre-processing, Machine Learning, Segmentation, Feature Extraction, CNN.

Abstract: Music plays a significant role in improving and elevating one’s mood as it is one of the important source of entertainment and inspiration to move forward. Recent studies have shown that humans respond as well as react to music in a very positive manner and that music has a high impact on human’s brain activity. Now a days, people often prefer to listen to music based on their moods and interests. This work focuses on a system that suggests songs to the users, based on their state of mind. In this system, computer vision components are used to determine the user’s emotion through facial expressions. Once the emotion is recognized, the system suggests a song for that emotion, saving a lot of time for a user over selecting and playing songs manually. A user study is presented to evaluate the application. Results reveal good usability of the service and show that visualization and interaction in a music emotion space can improve acceptance of music recommendations.

Abstract: Gender classification from facial photos is difficult due to the presence of a complex background, object occlusion, and varying lighting conditions. Face photos can be used for a variety of applications, including expression analysis, recognition, and tracking. This research investigates two deep learning-based approaches for gender classification using face photos. These approaches include CNN and Alex Net. Experiments were conducted to assess the effectiveness of both models in identifying male and female classes from facial photographs. The results indicate that both techniques were effective for gender classification. Additionally, a comparison study was carried out between these two models and a few well-known techniques for classifying gender. 

Keywords: gender classification, gender recognition, CNN, Alex Net, Deep learning

Abstract: Image captioning involves automatically describing images using words, attracting attention from researchers in natural language processing (NLP) and computer vision. Recent advancements primarily adopt an encoder-decoder framework, utilizing convolutional neural networks (CNNs) to extract image features and decoders to generate descriptions. Integration of attention mechanisms into this framework has notably improved performance. Leveraging the Transformer model, known for its effectiveness and efficiency in NLP tasks due to its attention mechanisms, we propose a novel approach combining CNNs and Transformers for image captioning. Our model utilizes a Transformer-Encoder to extract refined image feature representations, enabling the Transformer-Decoder to focus on pertinent image details when generating captions. Additionally, adaptive attention in the Transformer-Decoder determines the optimal utilization of image information during caption generation. Through extensive training on the Flickr8K_dataset, our model achieves an impressive 86.21% accuracy, demonstrating its efficancy and value in image captioning tasks.

Keywords: Image Caption, CNN, Deep Learning, Transformer, Attention mechanism,Flickr8k dataset.

Abstract: This  This project has an aim to control the horn volume in cities and also in restricted area as such schools, parks, hospitals, old age homes, college, government offices and in speed limited areas etc. Some peoples are driving vehicles in a high speed and create noise of horn of vehicle. So the police are not able to monitor all those things. Driver does control the speed of vehicle at places. This paper provides a way for how to control the speed of vehicle and control the noise of horn without harming others. This project has an aim to control noise of horn automatically. The speed of any vehicles will be detected using accelerometer if the speed of vehicle is exceed in honking zone then the buzzer can buzz and get alert to driver about . cities and also in restricted area as such schools, parks, hospitals, old age homes, college, government offices and in speed limited areas etc. Nowadays in a fast moving world all the peoples are not have self-control. Controls are taken automatically by the use of electronic system. In this project we use GPS for indicating the nearby honking zone. Speed is measured by the help of accelerometer in the vehicle. The controller compares the speed. If it exceeds the limited speed the pizzobuzzer buzz and alerts the driver and controls taken automatically by driver and when vehicle is near at honking zone the switch can automatically decrease the volume of horn if it on. In this way our smart honking zone for smart cities project will be perform.

Keywords: Speed control, inter-vehicular communication, honking, Arduino, switch, GPS, Pizzobuzzer, noise control.

Abstract:  Cloud computing has revolutionized the way businesses and individuals’ access and manage digital resources. This paradigm shift towards cloud-based solutions offers scalability and cost-efficiency but also brings forth a range of security concerns. This research delves into the multifaceted security challenges that emerge in the realm of cloud computing, aiming to provide a holistic understanding of these issues. It investigates data security, network security, identity and access management, and compliance as integral components of cloud security. The paper  identifies DDoS assaults, and risks associated with shared resources, with a detailed analysis of their implications. In addition to examining traditional security measures like encryption, authentication, and authorization, this research assesses contemporary security paradigms, including zero-trust security models and DevSecOps practices, within the context of cloud security. The importance of adhering to industry standards and regulations, such as GDPR and HIPAA, is emphasized. To offer practical insights, real-world case studies and examples of security breaches within cloud computing environments are presented. These case studies underline the real-world consequences of security lapses, both economically and reputationally. The research concludes by delineating best practices and strategies for mitigating security risks in cloud computing, such as adopting multi-layered security approaches, conducting routine security assessments, and investing in employee training. In an era where cloud computing is fundamental to modern IT infrastructure, understanding and addressing security issues is paramount. This comprehensive analysis serves as a valuable resource for cloud practitioners, security professionals, and policymakers, offering insights to fortify cloud environments in the face of evolving security threats.

Keywords: data security, network security, identity, access management, compliance, data breaches, insider attacks, DDoS assaults, shared resources, encryption, authentication, authorization, GDPR, HIPAA, zero-trust security models, multi-layered security approaches, employee training, shared responsibility model.

Abstract: Websites can enhance their security and protect against malicious Internet attacks by implementing CAPTCHA verification to distinguish between human users and automated bots. Text-based CAPTCHAs are commonly used as they are easy for humans to solve but challenging for machines to decipher. This research introduces a CNN model that utilizes binary images to recognize CAPTCHAs efficiently. The project involves creating an advanced Captcha Recognition System using deep learning on a Raspberry Pi. In real-time, the Raspberry Pi processes images with the help of OpenCV, applying the trained model to authenticate captchas. This innovative approach demonstrates the practical use of deep learning on edge devices, strengthening security through automated captcha verification and showcasing the potential for IoT security solutions in real-world applications. Key terms: Convolutional neural network; OpenCV; Automated CAPTCHA verification.

Abstract: This research presents a detailed investigation of the manner in which state-of-the-art computer vision technologies are employed to assist agricultural landscape monitoring and vegetative area analysis. The approach that has been presented comprises of a multi-step procedure. Before applying complex feature extraction algorithms, the procedure begins with significant picture processing to assure the correctness of the data. Subsequently, diverse agricultural characteristics are separated using segmentation methods, and important traits are grouped together using clustering algorithms. In order to assess these solutions' performance and establish their resilience in aiding with appropriate farm management practices and boosting environmental monitoring abilities, rigorous testing is undertaken. This research greatly enhances the topic of eco-agriculture by presenting new ideas and viable solutions that employ computer infrastructures given by cloud computing.

Keywords: Farm monitoring; vegetation analysis; computer vision; image preprocessing; feature extraction; clustering; segmentation; Cloud computing; Remote sensing; Precision agriculture; Data analytics; Machine learning; Geospatial analysis; Satellite imagery; Environmental monitoring; IoT (Internet of Things); Big data; Digital agriculture; Sustainable farming; Land use classification; Crop health assessment; Spatial data processing.

Abstract: The key issue for the railway department has been to examine and monitor railway bridges, as urbanization expands, the availability of railways grows, and the railway system has greatly expanded throughout the nation. The expense of maintaining railroad bridges and associated costs with personnel have been a burden on the railroads. To ensure transportation safety, concrete bridge crack detection is critical. Deep learning technology has made it possible to automatically and accurately detect faults in bridges. The present methods are not accurate and they require a large size of dataset for model training and they require a high computational power model training. The proposed model is a convolutional neural network (CNN) based end-to-end crack detection model. The proposed model achieved a 95% detection accuracy.

Keywords: Deep Learning, CNN, Remote sensing, OpenCV, Keras, TensorFlow.

Abstract: In the discipline of informatics, artificial intelligence (AI) uses algorithms to process data and constantly refines its reasoning. AI, which began in the 1950s and has since evolved into "machine learning algorithms," encompasses Deep Learning for pattern recognition in medical images and Machine Learning for data analysis. With the use of augmented reality and virtual reality, AI has the potential to dramatically improve healthcare, especially for radiologists working in diagnostic imaging and interventional radiology. Working in diagnostic medicine and interventional radiologists. AI applications in the field of interventional radiology include patient selection, treatment planning, and training. Thorough research and validation are crucial to successfully integrating AI to improve patient care. This study examines the prognostic value of haemorrhagic transformation (HT) in acute ischemic stroke inferred from MRI-derived permeability measurements using MR perfusion images that was done.

Keywords: Artificial Intelligence (AI), Informatics, Data Processing, Hemorrhagic Transformation (HT), Acute Ischemic Stroke, Patient Outcomes, Predictive Model.

Abstract: Agriculture is the foundation of many large economies, like India and Maharashtra. The newcomers in the agricultural sector face the challenge of deciding which crops to cultivate on their farms. This problem needs addressing, and we are taking steps to solve it. To address this, we have developed a system that predicts suitable crops for farmers based on the natural content present in the soil and another parameters like weather, rainfall, humidity and many more. This solution aims to reduce farmers' losses and enhance production. Unlike existing systems, which are not fully functional and cannot effectively guide farmers in selecting crops, ours utilizes classification and regression algorithms for crop prediction. Our system can be used by the farmer’s on the web as well as the android phone’s as well. Proposed system uses a dataset which contains the samples of the crops with the required nutrition’s such as Potassium. Phosphorus, Nitrogen, pH, Humidity, Rainfall and many more features. We are using the K-Nearest Neighbor (KNN) machine learning algorithm which is Supervised Learning algorithm used for the classification and regression. System uses a pickle library of python to create the Machine Learning Model which we actually recommend the crop to be cultivate. Model then takes the input as the ingredient in the soil as a parameter and then KNN finds the best suitable crop for that particular type of soil.

Keywords: Machine Learning, Crop Prediction, K-Nearest Neighbour, Crop Recommendation, Classification, Regression, Machine Learning Model, Supervised Algorithm.

Abstract: Hand key point detection is crucial for facilitating natural human-computer interactions. However, this task is highly challenging due to the intricate variations stemming from complex articulations, diverse viewpoints, self-similar parts, significant self-occlusions, as well as variations in shapes and sizes. To address these challenges, the thesis proposes several innovative contributions. Firstly, it introduces a novel approach employing a multi-camera system to train precise detectors for key points, particularly those susceptible to occlusion, such as the hand joints. This methodology, termed multiview bootstrapping, begins with an initial key point detector generating noisy labels across multiple hand views. Subsequently, these noisy detections undergo triangulation in 3D utilizing Multiview geometry or are identified as outliers. These triangulations, upon re-projection, serve as new labeled training data to refine the detector. This iterative process iterates, yielding additional labeled data with each iteration. The thesis also presents an analytical derivation establishing the minimum number of views necessary to achieve predetermined true and false-positive rates for a given detector. This methodology is further employed to train a hand key point detector tailored for single images. The resultant detector operates in real-time on RGB images and exhibits accuracy on par with methods utilizing depth sensors. Leveraging a single-view detector triangulated over multiple perspectives enables markerless 3D hand motion capture, even amidst complex object interactions. Keywords: Convolutional Neural Network, Key point detector, Density Network with a Single Gaussian Model, Mixture Density Network, Degree of Freedom.

Abstract: In the rapidly evolving landscape of information security, steganalysis algorithms play a pivotal role in safeguarding digital content integrity. Three notable algorithms, JUNIWARD, JMIPOD, and UERD, stand at the forefront of this endeavour, each offering unique capabilities in detecting covert information embedding. JUNIWARD employs advanced statistical modelling and machine learning techniques to discern characteristic artifacts induced by popular data hiding methods. This results in high detection rate while maintaining a low false positive rateS, solidifying its position as a significant advancement in steganalysis technology. JMIPOD, tailored for JPEG- compressed images, leverages sophisticated feature extraction and statistical analysis to identify subtle discrepancies introduced by covert information embedding. By exploiting vulnerabilities in the JPEG compression process, JMIPOD achieves impressive detection rates across a wide range of embedding rates, ensuring the integrity of digitally compressed content. UERD, the Universal Ensemble for Robust Detection, presents a pioneering approach by employing an ensemble of carefully curated classifiers. This methodology capitalizes on the complementary strengths of multiple steganalysis methods, leading to enhanced robustness against a broad spectrum of steganographic schemes. Rigorous experimentation across various datasets showcases UERD's superiority in detection performance and adaptability to evolving data hiding methodologies

Keywords: Steganalysis, Information Security, Digital Content, Integrity, JUNIWARD, JMIPOD, UERD.

Abstract: This innovative project leverages automation to streamline library management [1]. A robot, guided by a pre-established path, tracks shelf layouts in the library. Controlled by an Arduino UNO, it receives input from a PC regarding the desired book. Upon reaching the designated shelf, the robot compares saved RFID tags with the books present. Once the match is found, its robotic arm retrieves the book [2]. Remarkably, it can also return books to their shelves autonomously. This integration of robotics and technology not only enhances efficiency but also minimizes manual labour in library operations.

Keywords: Arduino UNO, RFID, Robotic arm, Automation.

Abstract: Face detection has been around for ages. Taking a step forward, human emotion displayed by face and felt by brain, captured in either video, electric signal (EEG) or image form can be approximated. Emotion recognition is a critical aspect of human-computer interaction, enabling machines to understand and respond to human emotions effectively. This research focuses on the development and implementation of a real-time image processing system for emotion recognition. The objective is to create an efficient and accurate system capable of recognizing facial expressions in real- time, paving the way for applications in diverse fields such as human-computer interaction, healthcare, and entertainment. The proposed system leverages advanced image processing techniques, including facial feature extraction, machine learning algorithms, and real-time data processing to analyze facial expressions accurately. A dataset comprising diverse facial expressions is used to train and validate the system, ensuring robust performance across a wide range of emotional states. The research also explores the integration of deep learning models, such as convolutional neural networks (CNNs), to enhance the system's ability to discern subtle nuances in facial expressions. To achieve real-time processing capabilities, parallel computing and optimization techniques are employed to streamline the computational workload. The system is designed to operate seamlessly on resource-constrained devices, making it applicable to a variety of platforms, including mobile devices and embedded systems Index Terms: Real time Image, electric signal (EEG), convolutional neural networks (CNNs).

Abstract: The Ambulance Pulse Risk Prediction system proposes an AI-driven proactive emergency response solution for hospitals, aiming to revolutionize traditional reactive measures by predicting medical emergencies before they occur. Leveraging real-time data from diverse sources including patient health records, wearable devices, and environmental factors, advanced machine learning algorithms analyze patterns and correlations to identify heightened risks such as cardiac events and strokes. Through proactive alerting, hospitals can allocate resources more efficiently and intervene preemptively, potentially preventing emergencies and improving patient outcomes. Key components encompass data gathering, algorithm development, system integration, and validation with a strong emphasis on privacy and ethical considerations. By harnessing the power of AI, this system has the potential to transform emergency medical services, enhancing patient care and saving lives on a global scale.

Keywords: Pulse sensor, Artificial intelligence, IoT, Ambulance Pulse Risk Prediction, CatBoost Algorithm.

Abstract: This term paper presents an imaginative approach to improve DevOps robotization through the integration of notices and manual endorsement workflows inside Amazon Web Administrations (AWS) Cloud Advancement Unit (CDK) Pipelines. The essential objective of this think about is to supply a comprehensive understanding of how AWS CDK Pipelines can be designed to consistently join notices and manual endorsement stages, subsequently streamlining the CI/CD process while guaranteeing straightforwardness and control. To achieve this goal, a straightforward technique was used, which included an examination of AWS CDK Pipelines' capabilities, the identification of applicable notification instruments, and the execution of human endorsement procedures. Real-world case studies and instances were analysed to determine the impact of these factors on arrangement speed, consistency, and client fulfilment. The key findings of this study reveal that incorporating notices and manual approval workflows into AWS CDK Pipelines not only improves the perceivability of arrangement forms, but additionally enables organisations to strike a balance between mechanisation and human intervention, adjusting to security and compliance requirements. Organisations may effectively caution partners and ensure that fundamental choices are taken amid the CI/CD pipeline execution by using administrations like Amazon SNS and AWS Step Capacities. This investigation's promises rest in its arrangement of key experiences and practical guidance for DevOps experts and AWS CDK clients. To encourage the use of these technologies, the article provides a step-by-step implementation guide, best practises, and real-world use examples. Furthermore, it emphasises the potential for improving existing CI/CD pipelines and sets the groundwork for future research in the evolving topic of DevOps computerization inside the AWS environment.

Keywords: AWS CDK Pipelines, DevOps Automation, Notifications, Manual Approval, Continuous Integration (CI), Continuous Deployment (CD)

Abstract: The modern era of this digital age has a lot of importance for email communication for personal and professional purposes. But this has been a source for scams and cyber based threats where attackers use spam emails and send malware virus based emails where if a user opens the mail the viruses affect the device the user is using because of which the cyber attacker can access and manipulate your data for their personal use and benefits . So to avoid all these kind off outbreaks in this paper we are going to describe how can we filter these kind of emails to avoid the user from attempting to open these and become a victim to this scam ,this can be done using machine learning techniques like Support Vector Machine, Naive Bayes Classifier,Decision Trees which help classify these emails as spam and ham which means the emails of proper content and with malicious content are categorized based on which cyber scams can be prevented. The creation of these kind of effective machine learning filtering systems are vital for prevention of scams present in this era because these are the kind of systems which can help prevent cyber scams we have even used classification techniques and deep learning based approaches like Artificial Neural Network which help improve accuracy and robustness in the spam email classification and methods like Random Forest and KNN are used for ensembling whereas NLP(Natural language Processing ) processes email text while keu attributes for classification are extracted by feature engineering. this approach involves a combination of various machine learning methods and efficiently trained models with proficient data for spam email detection.

Keywords: Email spam, SVM ,KNN,Random Forest,Decision Trees machine learning,NLP,Naive Bayes, cybersecurity, data manupulation, filtering techniques, email scam.

Abstract: This paper introduces a novel approach to text extraction and conversion using PyPDF2 technology, aimed at enhancing literacy education. A web application is developed to facilitate the conversion of images to Excel format, with a focus on leveraging PyPDF2 functionalities. The research investigates various methodologies for image-to-text conversion, highlighting the advantages and challenges associated with PyPDF2 compared to traditional OCR techniques. By addressing identified gaps in existing literature, the study presents a comprehensive methodology consisting of capturing, extracting, recognizing, and convert ing phases within the web application. Unlike conventional OCR methods, PyPDF2 offers improved text processing and segmentation algorithms, resulting in enhanced accuracy and efficiency in text extraction. The web application seamlessly converts uploaded images into editable text, making it a valuable resource for both literacy education and teaching staff in diverse educational settings.

Keywords: Recognition; PyPDF2: Text Extraction

Abstract: Flight delays are a significant concern for both passengers and airlines, leading to inconvenience, financial losses, and operational disruptions. This abstract presents a comprehensive approach to mitigating flight delays through the development of a full-stack web application leveraging big data and machine learning techniques. The proposed system utilizes a vast array of data sources, including historical flight data, weather conditions, air traffic, airport congestion, and aircraft maintenance records. By integrating and analysing these diverse datasets, the application aims to identify patterns and correlations that contribute to flight delays. Machine learning algorithms play a pivotal role in predicting flight delays accurately. Through the application of supervised learning techniques such as regression, classification, and ensemble methods, the system learns from historical data to forecast the likelihood of delays for future flights. Additionally, advanced models capable of handling complex relationships and nonlinearities are employed to enhance prediction accuracy. The full-stack architecture of the web application encompasses both front-end and back-end components, ensuring a seamless user experience. The front-end interface provides users with intuitive features for inputting flight details, accessing delay predictions, and receiving real-time updates. Meanwhile, the back-end infrastructure manages data processing, model training, and prediction generation in a scalable and efficient manner. The Flight Delay Prediction Full Stack Web Application represents a comprehensive solution for addressing flight delays through the synergistic integration of big data and machine learning technologies. By empowering stakeholders with timely and accurate predictions, the application has the potential to significantly mitigate the impact of flight delays on both passengers and airlines alike.

Keywords: Big data, Machine learning, Regression, Full stack architecture.

Keywords: DevOps, Jenkins, Automation, Continuous Integration, Continuous Delivery, Software Development Lifecycle.

Abstract: An innovative approach called "Cipher Safe: We intend to unveil on the market the "Digital Password Guardian" product which was created to make strong the password protection of internet services. It is a highly efficient method, a much-needed measure in our data insecurity system. It offers the simplest, yet the most effective way of dealing with preventing password compromise and any further unwanted access. Cipher Safe is as it can be a total mender of the current security woes that are caused by shortcomings such as interruption of the operation and delay in identification through biometric authentication as well as key exchanging that are on the updated mode. Participation in the program empowers people to train and have a clear understanding of the key password practices which is beneficial in that they can guard their identities more safely by having online portfolios too. By being the leaders in the technology averse to the threats, we are always at the tricky part of safeguarding ourselves against such vulnerabilities that are common thing in the bitcoin industry. This is because we continue updating and upgrading our software as soon as the threats are detected. The term does not provide any private user information, even if it complies with the best practices security protocols by employing encryption mechanisms which is presented with strong encryption; therefore, it could save any information regarding the user. It does not matter what platform users are utilizing - all the technologies are symmetric thanks to the Cipher Safe that ensures interoperability, and thereby cybersecurity. With Cipher Safe, over and above the function of a typical password manager, you will experience enhanced safe session gentrification and a broader scope of overall security.

Keywords: Cipher Safe, Encryption

Abstract: In developing countries, maintaining roads is one of the most crucial difficulties. Well-maintained roads contribute greatly to the nation's economy. To ensure safe driving, it is therefore essential to recognize potholes and determine their depth. An Arduino UNO, an ultrasonic sensor, servo motors, DC motors, and other electronic parts are used in the design and creation of the model. In order to do this operation, sensors that identify potholes and transmit the signal to the Arduino UNO are used. The pothole on the road is then fully filled after the signal is sent to the circuit that powers the different motors and dispenses the required amount of cement into the identified hole. The cement is then levelled using a roller.  

Keywords:  Well-Maintained roadways, Potholes, Dispense, Arduino UNO.

Abstract: The hexapod robot is one of the important classes in legged robots due to its great potential to operate in complex situation with high stability and flexibility. The  robot is designed by using Arduino as its central control unit, integrating a laptop camera for visual data acquisition, Zigbee for communication, IR Sensor and Metal Sensor for obstacle detection, Missile Identification for threat assessment, Person Face Authentication for access control, and a Laser Gun for precise target engagement. The proposed system enhances the security in defense , by identifying threats and functions in challenging environments by safe navigation. Key terms: hexapod robot, defense, zigbee, ir sensor, metal sensor, laser gun.

Abstract: The digital sector is filled with numerous platforms that enable online communication. Apart from interruptions due to spam messages and security breaches, these have posed a significant challenge. The purpose of this paper is to provide a comprehensive way of dealing with the issue of spam on chat platforms. We did so by using machine learning algorithms such as Naive Bayes, Support Vector Machine (SVM), Logistic Regression, et cetera in order to see how effectively they can be used in identifying and filtering out spam messages.The study involved an extensive comparison between accuracy and precision metrics for evaluating the performance of these algorithms.These algorithm’s strengths and limitations are shown through experimentation and analysis that give clues into what to consider when developing algorithms for detecting spam messages in various online chat platforms.In this light, we have successfully applied Naïve Bayes and logistic regression to text based Spam classification. After conducting thorough tests on it, the system had been able to identify 97% of all spams accurately resulting into a Precision Score of 1 thus enhancing trustworthiness and safety measures of online communication portals. Precision is preferred over accuracy due to imbalance in data.

Keywords: Text Spam, Naive Bayes, Logistic Regression, Chats, Spam Detection

Abstract: In the ever-evolving industrial landscape, worker safety remains a paramount Concern. With the advent of Industry 4.0 and the proliferation of connected devices, the Internet of Things (IoT) has emerged as a transformative force in enhancing workplace safety. IoT-based wearable systems offer a promising solution to realtime monitoring and Intervention, minimizing workplacehazards and fostering a culture of safety. Traditional safety Measures often rely on passive approaches, such as personal protective equipment (PPE) and Safety training. While essential, these methods may not always prevent accidents or provide Timely intervention in critical situations. IoT-based wearable systems address these limitations By transforming workers into active participants in their own safety. These systems compriseWearable devices equipped with a suite of sensors that collect real-time data on the worker’s Environment and physiological parameters. This data is then transmitted wirelessly to a Central hub foranalysis andvisualization. By continuously monitoring workers’ exposure to Hazards, such as toxic gases, excessive noise, or extreme temperatures, IoT-based wearable Systems can trigger immediate alerts and initiate preventive measures of fatigue, heat stress, or Potential health risks, enabling early intervention and preventing accidents.

Abstract: This project introduces an innovative IoT and ML-based smart healthcare monitoring system, integrating a smart bed with a light sensor, to track sleep patterns and environmental conditions. The system collects data on sleep quality and ambient light levels, facilitating detailed analysis and insights into patient well-being. Leveraging machine learning algorithms, it offers predictive analytics for proactive care and clinical decision support for healthcare providers. Through a user-friendly interface, patients engage with their healthcare data, enhancing awareness and adherence to treatment plans. The system's streamlined monitoring processes optimize resource allocation and contribute to improved healthcare outcomes and cost efficiency. Overall, this project showcases the potential of advanced technologies to revolutionize patient care delivery and promote personalized healthcare management. 

Keywords: healthcare, sleep pattern, sensors, patient

Abstract: The project addresses the crucial issue of drowsy driving, a significant contributor to road accidents. The proposed system employs cutting-edge technology to monitor the driver's level of drowsiness through computer vision. By analyzing facial cues and eye movements, the system can detect signs of fatigue and alertness in real-time. When drowsiness is detected, the system takes proactive measures to keep the driver awake and alert. It automatically plays an alarm to stimulate the driver's senses and mitigate drowsiness. The system also includes a hand detection module to determine whether the driver is active if the alert is activated a certain number of times. This function is essential since inactivity during a state of exhaustion might result in serious consequences. The technology alerts the driver and averts possible collisions when it detects a driver's condition of drowsiness. The technology reacts instantly in emergency scenarios if the driver's fatigue approaches a threshold and prolonged inactivity is noted. The driver's current position is shared with the chosen emergency contacts through an alert message. The overall goal of this project is to improve road safety through the efficient detection and mitigation of driver fatigue and the provision of prompt assistance in emergency situations.

Keywords: Computer Vision, Drowsy driving, Alarm, Alert message, Eye and hand detection.

Keywords: Command Line, Generative AI,Command Prompt

Abstract: The integration of IoT, automation, and advanced technologies such as artificial intelligence (AI) and deep learning has sparked a significant transformation in modern agriculture. In particular, the utilization of deep learning techniques, notably convolutional neural networks (CNNs), has emerged as a promising approach for disease detection in crops. This paper presents a comprehensive review of recent advancements in using deep learning, specifically focusing on the application of convolutional neural networks like VGG-16 in identifying plant diseases from leaf images. By harnessing the power of deep learning and leveraging tools like PyTorch, this study aims to revolutionize disease identification processes in agriculture. The automatic learning and feature extraction capabilities of deep learning offer a more objective and efficient means of detecting plant diseases compared to traditional methods. Moreover, the implementation of deep learning in agricultural settings promises faster and more accurate detection, enabling timely interventions for disease management. The review also addresses current challenges and future directions in the field, providing valuable insights for researchers and practitioners working on disease detection and pest control in agriculture.

Keywords: - Convolutional neural network, Deep neural networks, Deep structured learning, Machine learning.

Abstract: The "Street Veggie" project is all about making things better for people who sell food on the street. We're doing this by creating a special android application just for them. Selling food on the street is important in cities because it provides affordable and easy-to-get food for lots of people. But, there are challenges like not being seen enough, worries about cleanliness, and some problems in how they run their businesses. "Street Veggie" is here to help fix these issues by using technology. Our application is like a helpful tool for street hawkers. It lets them show all the different kinds of food they sell and reach more customers using their phones. This way, people can easily find and connect with street hawkers. The app also has easy tools to help hawkers manage their food, keep track of sales, and set the right prices. These tools make it simpler for them to run their businesses well and earn more money. "Street Veggie" isn't just good for individual hawkers; it also makes street food in cities more exciting and lively. This paper talks about how our project works and shows how using technology can make life better for street hawkers, helping them earn money while making the street food scene in cities even more enjoyable. "Street Veggie" is an exciting project that wants to make a positive change in how street food is sold, making it more modern and better for everyone involved.

Keywords: street hawkers, street veggies, application, customers, business.

Abstract: Heart disease is one of the most common causes of illness and mortality worldwide, along with other cardiovascular disorders. Early detection and diagnosis of heart disease are crucial for preventing serious complications and saving lives. The study described in this abstract, "Advanced Predictive Models for Early Heart Disease Detection: Harnessing Embedded Machine Learning," investigates cutting-edge machine learning methods incorporated within medical applications to improve the early identification of heart disease. The study focuses on the employment of embedded machine learning algorithms in a broad framework created to analyse many health-related data sources, including electronic health records, wearable device data, and medical imaging. These artificial intelligence (AI) models are integrated into the healthcare infrastructure to provide real-time data analysis and prediction, enabling the early detection of those at risk for heart disease. The study emphasises the important benefits of embedded machine learning, including scalability, real-time tracking, and seamless integration with healthcare systems. Additionally, it talks on the difficulties with data privacy, data quality, and model interpretability when applied to embedded machine learning for the early diagnosis of heart disease. The findings of this study show the potential to fundamentally alter the prognosis of heart disease, thereby easing the burden of this serious health problem. This ground-breaking method provides the path for the creation of more individualised, precise, and effective instruments for the early detection and control of cardiac disease.

Keywords: Heart disease, Predictive models, Early detection, Embedded machine learning, Healthcare technology, Cardiovascular risk assessment

Keywords: Health monitoring, Voice-voice-interaction, Emotional support

Abstract: Non-invasive vein detection for intravenous(IV) procedures can be carried out using infrared(IR) rays for the purpose of illuminating a region, and then using an infrared camera for observing it. The image is processed using different techniques on the Open-CV software platform. We further propose the model for patient health parameter reading like BPM, SPO2, Temperature on google Firebase cloud. A This project works on the principle of absorbance of infrared (IR) light by veins and its diffusion to surrounding tissues which makes the vein appear darker when viewed through an IR sensitive camera.One of the challenges faced while implementing this system is to make an efficient system for image acquisition and image processing at low cost. Thus, a customized sensitive camera is used to cut down the cost.

Keywords: Intravenous, infrared rays, open-CV, image acquisition.

Abstract: This research paper presents the development and implementation of the "Farm" Android application, designed to facilitate online fertilizer selling and accounting. The application serves as a comprehensive platform aimed at streamlining the process of fertilizer procurement for farmers while providing efficient inventory management and accounting features for suppliers. Through a user-friendly interface, "Farm" offers functionalities tailored for both users and administrators. For users, "Farm" provides essential features such as user authentication, profile management, product browsing, cart management, order placement, and product filtering. Users can create and manage their profiles, browse through a diverse catalog of fertilizers, add products to their cart, place orders securely, and apply filters to refine their product search. Additionally, a forgot password feature ensures seamless user access in case of forgotten credentials. Administrators, on the other hand, benefit from an array of management tools to oversee product listings, orders, and inventory. Admin functionalities include login authentication, product management (addition, editing, deletion), order viewing, and inventory management. Admins can easily add new products, edit existing listings, view past orders, and manage inventory levels to ensure smooth operations. The development process leverages technologies such as Android Studio for front-end development, Firebase for back-end and cloud services, and Java for programming logic. The application architecture follows best practices to ensure scalability, reliability, and security. Key features are implemented using Firebase Authentication for user authentication, Firebase Realtime Database for data storage, and Firebase Cloud Messaging for push notifications.  

Keywords: Android application, fertilizer selling, accounting, online marketplace, user authentication, product management, inventory management, user experience, Firebase, Java programming, agricultural supply chain.

Abstract: Businesses, organizations, and educational institutions all require reliable and accurate attendance monitoring solutions in today's hectic world. This paper presents an automated attendance system employing facial recognition technology based on CCTV video. The system uses strategically placed CCTV cameras to capture real-time video footage of individuals entering or leaving the premises. Facial detection algorithms locate and identify human faces, and facial recognition algorithms extract unique facial features to create facial templates. These templates are compared to a pre-registered database for facial matching, and attendance is recorded upon successful identification. Advanced features like liveness detection enhance accuracy and security.Attendance data is securely stored in a centralized database and can be integrated with other systems. This study describes an automatic attendance system that uses CCTV video-based facial recognition technology.

Keywords: facial recognition, automatic attendance system, CCTV, centralized database.

Abstract: This project aims Breast cancer remains a significant global health challenge, necessitating advancements in predictive modeling to enable early detection and personalized treatment. Traditional centralized machine learning approaches often face privacy concerns, especially when dealing with sensitive medical data. Federated Learning (FL) emerges as a promising solution, allowing model training across decentralized devices without sharing raw data. This project explores the implementation of Federated Learning for Breast Cancer Prediction, aiming to improve both privacy and prediction accuracy. The federated learning framework involves collaboration among multiple healthcare institutions, each possessing a subset of breast cancer patient data. The model is trained locally on each institution's data, and explored for decentralized model training on distributed data sources, preserving privacy and security. The trained federated model is evaluated and saved for deployment. Further research in this area holds promise for revolutionizing healthcare delivery worldwide.

Keywords: Breast cancer prediction, Artificial intelligence, Deep learning, Centralized learning, Decentralized learning, Image-based analysis, Machine learning, Healthcare applications.

Abstract: Timely diagnosis of diseases is considered vital for treatment. So far, many methods have been created for this purpose; autism spectrum disorder is one of these diseases. In this disease, developmental and developmental disorders will accompany the patient especially since childhood. The range of this disease is wide and it can usually be diagnosed by performing a series of clinical tests. Usually, the symptoms of this disease appear in childhood and before the age of three, and they differ according to the severity of the disease and its symptoms, which may appear in some cases from 5 months of age to two years of age. One of the important and debatable points is the timely diagnosis of this disease in adults, which unfortunately was not diagnosed in childhood, and this causes a series of behavioral problems in the social life of people with autism, and the person, by referring to a neurologist and Nerves and performing a series of clinical evaluations are known to be suspicious of autism spectrum disorder. Therefore, it is important to provide methods that can identify the relationship between different characteristics in contracting this disease. In this research, an attempt is made to predict the effect of each of the parameters on the diagnosis of the disease and also the process of the disease by using Bayes' law and genetic algorithm. In this method, mutual validation technique is used to optimize input and output data. First, the data are pre-processed and in the next step they are classified by Naive Bayes (kernel) which achieves 91% accuracy and then they are optimized with genetic algorithm which reaches 94.06% accuracy. Also, the data were tested with decision tree and Naive Bayes algorithms, and their results were compared. Keywords: autism, Bayes law, Naive Bayes, Naive Bayes (Kernel) genetic algorithm, cross validation Cite: Mohammadali Mohammadi, "Diagnosis of Autism Spectrum Disorder in Adults by Combining Bayes' Law and Genetic Algorithm", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 4, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.134104.

Keywords: Online rental Framework, Online rental things, Environment awareness.

Abstract: Railways place a premium on passenger safety. The goal of this study is to build a control at platform while the passengers crossing the platform with automation operation m response to the arrival and departure of trains. As a tram approaches the crossing, the gate closes automatically. Two infrared sensors monitor the tracks and alert the operator when a train approaches or departs. Motors, controlled by an Arduino Uno, will operate the gate's opening and shutting. The closure of the gate is signalled to those waiting to pass using buzzers. The growing frequency of accidents at platforms in India may be mitigated with the use of this method. Arduino C code provides hardware support. There is less waste and greater safety in the suggested approach.

Keywords:  Paraphrase recognition, passage-level plagiarism detection, support vector machine.

Abstract: Around the world, thousands of individuals with hearing loss use sign languages which come in a variety of regional variations to communicate daily. Therefore, it is believed that strengthening communication and inclusion for this group of people requires the automated translation of sign languages. But this is challenging research subject due to a number of issues. Developing a consistent system is impractical due to the regional variations in sign language. This is one of the key obstacles. Still, sign language recognition technology holds promise for improving deaf community services by assisting in communication gaps and enhancing the general well-being of society. The goal of the” Real-Time Hand Gesture Detection for Sign Language Recognition using Python” project is to create a system that can instantly translate sign language movements into text. The system tracks and detects hand movements using computer vision techniques, and then classifies the gestures using machine learning algorithms. The classes” Ok, Open Hand, Peace, thumbs up Thumbs down and alphabets A-Z” can be detected by our suggested system. Python programming and the OpenCV library will be used to carry out the project’s computer vision tasks. In our suggested system, we create two distinct sections: the first uses the Exception architecture model to recognize hand motion photographs and forecast outcomes, while the second uses OpenCV to detect in real time using a webcam. With the help of the Exception architectural model, our suggested solution was able to achieve 90.34% training accuracy and 90.00% validation accuracy. A dataset of hand motions that were recorded with a webcam will be used to train the hand gesture identification model. The finished system will include an intuitive user interface to let people who don’t use sign language communicates with others who do. In a variety of contexts, including public areas, workplaces, and classrooms, it may enhance inclusivity and communication for individuals with speech or hearing impairments.

Keywords: Hand Gesture, Segmentation, Random Forest, OpenCV, TensorFlow, Keras, CNN, Deep learning, Scikit learn.

Abstract: The AI-based interview evaluator is a comprehensive system designed to provide an objective and data-driven assessment of candidates during job interviews. By leveraging cutting-edge technologies in machine learning, computer vision, and natural language processing, the system analyses video and audio inputs to evaluate a candidate's emotions, confidence, and knowledge. For emotion recognition, the system utilizes Deep face and Haar Cascade models, which can detect a wide range of facial expressions and subtle emotional cues. These models help determine the candidate's emotional state throughout the interview, providing valuable insights into their level of engagement and comfort. In addition, the system employs Google Speech Recognition for accurate speech-to-text conversion, allowing it to analyse the content of the candidate's responses. This feature enables the system to assess the candidate's communication skills, articulation, and knowledge of the subject matter. To evaluate the candidate's confidence levels, the system utilizes a Random Forest Classifier trained on datasets containing confident and non-confident speech patterns. By comparing the candidate's speech patterns against these datasets, the system can determine their level of confidence in their responses. Furthermore, a neural network-based chatbot is integrated into the system to provide a more interactive interview experience. The chatbot can ask follow-up questions, clarify doubts, and engage the candidate in a conversation, simulating a real-life interview scenario. Based on the analysis of the candidate's emotions, confidence, and knowledge, the system generates insights and suggestions to aid organizations in making informed hiring decisions. These insights can help identify candidates who are well-suited for the role and provide valuable feedback for candidates looking to improve their interview performance.

Keywords: AI Based Interview Evaluator, Facial Expression Analysis, Deep Face, Machine Learning, Convolution Neural Network, Speech-based Confidence Detection, Librosa, mfcc, confidence evaluator, AI interview.

Keywords: Adaptive Learning, AI-LMS (Artificial Intelligence – Learning Mechanism System), Personalized Learning, Intelligent Tutoring System, Virtual Assistant

Abstract: Today in the current global scenario, women are facing many problems like women’s Harassment. We propose to have a System which is the integration of multiple devices, Hardware comprises of Portable system that endlessly communicates with a sensible phone that has access to the web. This paper covers descriptive details about the design and implementation of” System”. The System consists of an A9G Module, GSM module (SIM900A), GPS module (Neo-6M), AG9Module (RDA8955), Bluetooth (HC-05) Module, Panic Button. In this project, when a woman senses danger she has to press the Panic Button of the device. Once the system is activated, it tracks the current location using GPS (Global Positioning System) and sends an emergency message using GSM (Global System for Mobile communication) to the registered mobile number and nearby police station. IoT module is used to track the location continuously and update it into the webpage with help of android safety app we can track location and ask for help with voice command and reaches to nearby contacts. The main advantage of this project is that this device can be carried everywhere since it is small and also provides safety to Women.

Keywords: A9G, GSM, GPS, Bluetooth module, Panic Button

Abstract: The rapid growth of online data transmission has increased the demand for stronger data security. Intrusion Detection Systems (IDS) are essential for identifying virtual security threats by using advanced technologies, especially Machine Learning Algorithms, to swiftly detect and categorize attacks in real-time and determine the most accurate algorithm for attack classification. The current setup uses various intrusion detection algorithms, with a focus on improving performance through advanced algorithms like the Ensemble Learning and Discriminate Analysis. Unlike the existing approach that primarily relies on accuracy, we have used performance parameters such as Accuracy, Precision, Recall, and F1-Measure for evaluating the performance of the models. This comprehensive analysis aims to improve intrusion detection, offering a deeper understanding of algorithm effectiveness, and increasing confidence in the system's intrusion detection capabilities.

Keywords: Machine Learning, Datasets, Feature Selection, Machine Learning algorithms, Intrusion Detection System

Abstract: Birds are an amazing creature which lead lovely lives along with humans which are one of the signs of Climatic change. Identification of Bird Species is a Complex Task for humans as there are huge number of species of birds are available. Even it is also more difficult for Ornithologists to identify the correct name of a bird related to a particular specie. The main importance of identifying the Bird Species includes various applications such as for monitoring wildlife, for the efforts of conservation from becoming extinct, and for some projects which are related to the birds. As the present existing system uses the Random Forest algorithm to identify the bird species from image. In proposed system tried to utilize deep learning algorithm models in order to enhance the overall accuracy of the project to identifying the bird image. We used EfficientnetB4 algorithm in order to increase the Accuracy.

Keywords: Random Forest, Decision Trees (DT), EfficientnetB4.

Abstract: Individuals with hearing impairments often encounter challenges in communicating effectively with those who do not share their condition. The majority of the population lacks awareness regarding the recognition of sign language. Employing machine learning and computer vision (CV) technologies can offer substantial support to the hearing impaired. These technologies can be further developed to create automatic interpreters, allowing individuals to comprehend sign language effortlessly through hand gesture recognition. In interpersonal communication, hand movements hold significant importance, serving as a crucial means to connect individuals with hearing impairments and those without.

Keywords: Hearing Impairments, Interpersonal communication, computer vision(CV), Hand Gesture Recognition

Abstract: The aim of the project is to develop an effective method for anomaly detection and recognition in live streams. Anomaly detection refers to the task of identifying instances or events that deviate from the expected or normal patterns. In the context of live streams, anomalies can include activities such as fighting, accidents, robbery, and other unusual events. Rather than requiring continuous human monitoring of surveillance feeds, automated anomaly detection can relieve the burden on security staff. The system can alert operators only when potentially significant anomalies are detected. Effective anomaly detection can help reduce false alarms in live streams. By accurately distinguishing between the normal and anomalous activities, security personnel can avoid unnecessary responses to non-threatening situations. The proposed method works better by recognizing the abnormal behaviors in long-distance captured images and reducing the false alarm rate by effectively distinguishing between the normal and abnormal activities.

Keywords: Anomaly detection, live streams, false alarms, unusual events, abnormal behaviors, normal activities, effective recognition 

Abstract: There have been significant strides in computer vision that result in momentous improvements in object detection and tracking, which form the basis of a number of applications such as surveillance, driverless vehicles and human-computer interaction. This paper proposes an original but complicated method for reliable and precise tracking based on DeepSORT (Deep Simple Online and Realtime Tracking) with YOLOv5 (You Only Look Once version 5). YOLOv5 is an effective detector that performs object detection by looking once on an image or video frame to identify objects as well as their locations. These detection results are then incorporated into the DeepSORT tracking framework, which employs deep learning techniques to consistently track objects across frames. The combination of YOLOv5 and DeepSORT addresses issues of accuracy in detecting as well as reliability in following objects thereby providing a holistic approach to dynamic scenes involving multiple objects. The proposed system detects many different yolov5s and DeepSORT at one time.

Keywords: YOLOv5, DeepSORT.

Abstract: Steganography stands as a prominent technique, allowing the covert embedding of secret data within seemingly innocuous files. This project introduces a comprehensive approach to multi-format steganography, enabling the concealment of data within images, audio, video, and text files. The system employs the efficient RC4 encryption algorithm, chosen for its simplicity, speed, and adaptability in various key sizes. Through a command-line interface and leveraging Python libraries such as NumPy, Pandas, OS, and Cv2, the project encompasses distinct modules catering to different file types. This research work bridges the gap between flexibility and security, offering a high performance, multi-format steganography tool. Its implementation not only demonstrates the practicality of data concealment across varied files but also sheds light on potential areas for further optimization and enhancement. This research fills a critical void by providing a comprehensive tool that balances flexibility and security, offering high-performance multiformat steganography capabilities. Its implementation not only showcases the practicality of concealing data across diverse file formats but also highlights potential avenues for further optimization and enhancement.

Keywords: steganography, images, audio, video, text files, RC4 encryption algorithm, speed, adaptability, Python libraries, NumPy, Pandas, OS, Cv2, flexibility, security.

Abstract: Textual vision, the fusion of natural language processing and computer vision, has gained significant attention in recent years due to its applications in tasks such as image captioning, text-based image retrieval, and visual question answering. In this paper, we explore the utilization of quantized latent spaces in textual vision tasks. Latent space representations, generated from textual data, capture semantic information essential for understanding and interpreting text. By quantizing these latent spaces, we aim to reduce dimensionality while preserving important semantic features. We present a methodology for generating quantized latent space representations from textual data and discuss the process of quantization using various techniques. Experimental results on benchmark datasets demonstrate the effectiveness of our approach compared to baseline methods. Our findings indicate that leveraging quantized latent spaces enhances the performance of textual vision tasks, paving the way for more efficient and interpretable text-based image processing systems.

Keywords: Quantized Latent Spaces, attention mechanism, VQ-VAE mechanism, Conditional GAN, computer vision (CV).

Abstract: Pall computing has surfaced as a revolutionary means of data sharing, enabling a vast number of individualities to pierce information over networks fleetly and efficiently. In surrounds similar as particular health record( PHR) systems, the traditional burden of carrying colorful paper documents for judgments has been soothed. rather, cases can upload their health records to PHR systems, granting them the capability to store, recoup, and widely partake their data with authorized parties, including musketeers, family, and healthcare providers. The necessity for precise access control in PHRs has urged a demand for encryption schemes able of enforcing fine- granulated access control. Retired ciphertext policy trait- grounded encryption( HCP- ABE) is a promising result to this challenge by cache access control programs within ciphertext, by enhancing sequestration protection. Unlike before mechanisms where access control programs were frequently transferred along with ciphertext, facing a threat to druggies' sequestration, HCP- ABE embeds the access structure directly into the ciphertext. This prevents unequivocal exposure of sensitive attributes similar as" cardiologist" or" central sanitarium" contained within access programs, which could unintentionally reveal a case's medical condition to unauthorized druggies. While the preface of HCP- ABE addresses sequestration enterprises, being schemes aren't without limitations. numerous of these schemes support only introductory sense gates similar as AND gates or combinations of positive, negative, and wildcard attributes. Accordingly, two significant downsides arise an increase in the size of public parameters commensurable to the number of attributes, and a substantial rise in decryption costs. To alleviate these issues, low- overhead schemes have been proposed. These schemes generally incorporate decryption tests involving the addition of spare factors to ciphertext before decryption. While this approach enhances decryption effectiveness, it also results in a significant increase in ciphertext length, potentially hindering overall performance. Sweats to optimize encryption schemes for PHRs must strike a balance between security and effectiveness. The exploration and development are necessary to address scalability and performance enterprises while icing strict access control and sequestration preservation. By these challenges, pall computing technologies can continue to transfigure data operation practices, particularly in sensitive disciplines like healthcare, fostering trust and confidence among druggies and easing wide relinquishment.  TERMS : Personal Health Records(PHR),Attribute-Based Encryption, Hidden Policy, Fast Decryption

Abstract: Defect detection has been revolutionized by the use of Convolutional Neural Networks (CNNs) for identifying defects in objects through image processing. While traditional CNN-based object detection algorithms have shown success in identifying natural objects, they often struggle when it comes to defect data. To tackle this challenge, a shared weight binary classification network is implemented to determine the presence of defects in images. This is then followed by a detection network that accurately locates the defects within the objects. By utilizing this approach, the speed and accuracy of defect detection are significantly improved compared to conventional CNN-based object detection methods. This has been supported both theoretically and experimentally, demonstrating the effectiveness of the shared weight binary classification network in enhancing defect detection using CNN technology.

Keywords: CNN (Convolutional Neural Network), Image processing, Defect detection, Object detection, Shared weight binary classification network.

Abstract: The proliferation of hate speech on social media has become a pressing societal concern, prompting the need for effective identification and mitigation strategies. This abstract outlines a novel approach utilizing machine learning (ML) and natural language processing (NLP) techniques to detect hate speech and assess its impact on inducing stress among users. The study focuses on the development of an ML-based model trained on a diverse dataset of social media content to accurately identify hate speech. Leveraging NLP, the model aims to comprehend linguistic nuances, context, and sentiment within textual data, enabling it to distinguish between normal discourse and potentially harmful language. Furthermore, the research extends beyond mere identification, aiming to gauge the psychological impact of hate speech by analyzing its correlation with stress levels among social media users. By employing sentiment analysis and stress identification algorithms, the study aims to quantify the emotional toll experienced by individuals exposed to such content. The abstract emphasizes the interdisciplinary nature of the research, bridging the gap between computer science, linguistics, and psychology. The proposed methodology holds promise in aiding social media platforms, policymakers, and mental health professionals in devising targeted interventions to combat hate speech and mitigate its adverse effects on users' well being. Through this holistic approach, this study endeavors to contribute to the development of proactive strategies for early detection, intervention, and support mechanisms, fostering a safer and healthier online environment for all users.

Keywords: Stressfull comments, hate speech, personal assaults, healthier online environment.

Abstract: Counterfeit currency remains a significant challenge worldwide, posing threats to economic stability and security. This paper presents a novel approach for detecting fake currency utilizing image processing techniques. The proposed system leverages the advancements in computer vision and machine learning to automatically identify counterfeit banknotes with high accuracy. Initially, the input currency image is preprocessed to enhance its quality and extract relevant features. Subsequently, feature extraction algorithms are applied to capture distinctive patterns and characteristics unique to genuine banknotes. These features are then fed into a machine learning model, such as a neural network or support vector machine, trained on a dataset comprising both genuine and counterfeit currency samples. Through extensive experimentation and validation, the effectiveness of the proposed method is demonstrated in accurately distinguishing between authentic and counterfeit banknotes. The system's robustness against various types of counterfeit techniques and its potential for real-time application make it a promising tool for combating counterfeit currency fraud. This research contributes to the ongoing efforts in developing reliable and efficient solutions for safeguarding financial systems against counterfeit threats. Keywords— Machine learning, Artificial learning, CNN and Image processing.

Abstract: The Soldier Health Monitoring System (SHMS) is a comprehensive solution designed to ensure the well-being of military personnel by continuously monitoring vital health parameters. This system incorporates a range of sensors, including a Heartbeat sensor, Temperature Sensor, Vibration Sensor, and GPS, interfaced with an Arduino microcontroller and Nodemcu for real-time data acquisition and analysis. The system provides valuable insights into a soldier's health status, allowing for timely intervention in case of anomalies. The core components of the system include a Heartbeat sensor for monitoring pulse rate, a Temperature Sensor for body temperature measurement, and a Vibration Sensor to detect external impacts or abnormal movements. These sensors are connected to an Arduino microcontroller, which processes the data and displays it on an LCD screen. Additionally, a GPS module is integrated to track the soldier's location, enhancing situational awareness. The Soldier Health Monitoring System features a relay and a Peltier device to regulate body temperature. In extreme environmental conditions, the Peltier device can be activated to either cool or heat the soldier's body, ensuring optimal physiological conditions. The relay also enables the triggering of alerts or alarms in emergency situations. To enhance communication and situational awareness, the system utilizes a Nodemcu module for wireless connectivity. In the event of abnormal health parameters or critical situations, the Soldier Health Monitoring System is programmed to send instant notifications to designated recipients through various communication channels, such as SMS or email. The combination of real-time health monitoring, environmental adaptation, and intelligent alert mechanisms makes the Soldier Health Monitoring System an effective tool for safeguarding the well-being of military personnel. This technology not only ensures prompt medical attention in emergencies but also facilitates proactive measures to optimize soldier performance and mission success.

Keywords: Microcontroller ATmega16a, solar panel, rechargeable battery, temperature sensor , heartbeat sensor, Peltier plate, GSM, GPS.

Keywords: - Precision Agriculture,Convolutional Neural Network (CNN),Leaf Disease Detection,Autonomous Agriculture Robot, Controlled pesticide spray.

Abstract: Building a secure electronic voting system that offers the fairness and privacy of current voting schemes, while providing the transparency and flexibility offered by electronic systems has been a challenge for a long time. In this work-in-progress paper, we evaluate an application of blockchain as a service to implement distributed electronic voting systems. The paper proposes a novel electronic voting system based on blockchain that addresses some of the limitations in existing systems and evaluates some of the popular blockchain frameworks for the purpose of constructing a blockchain-based e-voting system.

Keywords: blockchain; digital transformation; e-voting system; security; scalability; systematic review

Abstract: The goal is to create a user-friendly mobile app facilitating accurate translation between English and Hindi, encouraging extensive use of Hindi in official contexts. Bridging language gaps in official matters is crucial in our interconnected world, yet current translation tools often lack efficiency, leading to misunderstandings. Our solution entails developing a robust translation software with multi-input capabilities, leveraging state-of-the-art natural language processing techniques. The app will enable text, voice, and image inputs, aiming to simplify language learning and stress-free communication. Ultimately, our project seeks to break down language barriers, promote linguistic inclusivity, and enhance effective communication by empowering users to communicate effectively in their native language while embracing cross-lingual communication.

Keywords: Multi-input Language Translation, Optical Character Recognition(OCR), Firebase ML Kit, User-friendly Interface, Automatic Language Detection.

Abstract: Machine learning techniques have brought about a revolution in various fields, with a significant impact on cyber security. In the face of growing cyber threats, the need for effective intrusion detection systems (IDS) has become more crucial than ever. These systems play a vital role in the timely and automatic detection and classification of cyber attacks, at both the network-level and the host-level. However, traditional IDS, which rely on conventional machine learning methods, often fall short in terms of reliability and accuracy.As the number of network-related applications, programs, and services continues to grow, so do the associated network security issues. Safeguarding the network against malicious activities is a challenging and critical task. In order to maintain a secure network environment, an effective system for detecting and identifying any suspicious activity is essential. This system is commonly known as an Intrusion Detection System (IDS).

Abstract: Facial expression recognition plays a vital role in understanding human emotions and behavior. In the context of Autism spectrum disorder childrens, accurate recognition of their facial expressions holds significant potential for aiding diagnosis, treatment, and communication. This abstract presents an overview of a machine learning-based approach to facial expression recognition for children with the Autism Spectrum Disorder. They face challenges in effectively expressing their emotions verbally, which underscores the importance of nonverbal cues such as facial expressions. In proposed work, Machine learning algorithms will implement for acquiring better accuracy for recognizing facial expressions, offering a non-intrusive and objective way to assess emotional states. Bench mark dataset available online along with our own -prepared dataset containing images of Autism spectrum disorder childrens exhibiting a range of expressions will use for model training, testing and validation. Keyword: Autism in children; machine learning; Computer vision; convolution neural network (CNN)

Abstract: Eye Disease is a frequent health issue that can cause either partial or total loss of vision. Early diagnosis is essential for eye disease to be effectively treated and managed Deep learning has become a potent tool for the detection and diagnosis of numerous medical disorders, including eye diseases, in recent years. Convolutional Neural Network (CNN) is a deep learning technique that is often used for image analysis and pattern recognition. In this approach, we use a CNN based method for classifying various eye diseases. The suggested method extracts feature from eye images and categorizes them using CNN.A dataset of eye images. collected, including images of healthy eyes. well as diseased eyes which are Cataracts, Glaucoma, Bulging eyes, Uveitis, and Crossed Eyes. The dataset underwent preprocessing to normalize the images and get rid of any noise. The training and testing sets were then created from the pre-processed dataset. A CNN model is trained using the pre-processed images, and it is then tuned using one among the many fine-tuning methods. The proposed approach will achieve high accuracy in detecting eye conditions. The results indicate that this approach can be a valuable tool in the early detection and diagnosis of eye conditions, which can improve the outcomes of treatment and prevent vision loss.

Keywords: Convolutional neural networks, Cateract, Dataset, preprocessing.

Abstract: The website security is an important issue that must be pursued to protect Internet users. Traditionally, blacklists of malicious websites are maintained, but they do not help in the detection of new malicious websites. This work proposes a machine learning architecture for intelligent detecting malicious URLs. Forty-one features of malicious URLs are extracted from the data processes of domain, Alexa and obfuscation. ANOVA (Analysis of Variance) test and eXtreme Gradient Boost (eXtreme Gradient Boosting) algorithm are used to identify the 16 most important features. Finally, dataset is used to learn the eXtreme Gradient  Boost classifier, which has a detection accuracy of more than 98%. Keywords: eXtreme Gradient Boosting algorithm; Malicious URL;. Feature Analysis; Chrome Extension

Keywords: mobile application, violation, unified modelling language, data management

Abstract: The stock market, as a preferred path for investment, continues to attract a growing number of individuals. Yet, the attraction of potential profits is counterbalanced by the substantial risks associated with stock market investments. In response to this dynamic and complex financial environment, the field of machine learning has been harnessed to construct predictive models that offer insights into future stock price movements. Support Vector Regression (SVR) and Long-Short Term Memory (LSTM), two distinctive and effective machine learning techniques, are used in this research project to explore the field of stock price prediction. The closing values of the stocks of five different companies are forecasted using these models. The Root Mean Squared Error (RMSE), one of the most well-known error metrics, is carefully used to assess the predictive accuracy and performance of SVR and LSTM. The findings of this empirical study show a striking difference between the two methods. In this comparison examination, LSTM is shown to be the better option, demonstrating its outstanding abilities to capture the complex dynamics and nonlinear patterns present in stock price data. The study's conclusions help us comprehend the potential of machine learning for stock market forecasting by highlighting the advantages of LSTM as a stock price prediction tool. Keywords: Machine Learning, Stock Market, Stock Price Prediction, Artificial Neural Network, Recurrent Neural Network, Long Short-Term Memory, Support Vector Machine.

Abstract: Sports activity recognition plays a crucial role in various applications, including athlete performance analysis, sports broadcasting, and injury prevention. Traditional methods for activity detection often rely on manual observation or rule- based systems, which are labor-intensive and lack scalability. In recent years, deep learning algorithms, particularly convolutional neural networks (CNNs), have emerged as promising tools for automated sports activity detection. This research paper presents a comprehensive investigation into the application of deep learning techniques for sports activity detection. We propose a CNN-based model and evaluate its performance against existing methods using standard sports activity datasets. Our results demonstrate the effectiveness of the proposed approach in accurately detecting sports activities, surpassing traditional machine learning approaches and achieving competitive performance compared to state-of-the-art models. This study contributes to the advancement of sports analytics and provides valuable insights for researchers and practitioners in the field of activity recognition.

Keywords: Sports activity detection, Deep learning, Convolutional neural networks (CNNs), Performance evaluation, Sports analytics.

Abstract: In response to growing safety concerns and increasing accidents and traffic violations, our project offers an advanced solution that combines sensor-based technology, machine learning algorithms and GPS technology to create a powerful system to solve the accident and traffic problem. Violation of the law. The main aim of the project is to create an integrated system that will not only detect incidents in a timely manner, but also manage emergency responses and reduce their impact. The proposed system uses state-of-the-art sensors to continuously monitor multiple vehicles and can detect anomalies that indicate an accident. Using machine learning algorithms, the system analyses this data to distinguish between driving behavior and situations that require intervention. The addition of GPS technology increases the accuracy of the system, provides accurate location information, and provides instant alerts to drivers and authorities. The general purpose of the project is not only accident research; It aims to contribute positively to improving road safety, thus saving lives, and reducing industrial accidents and damage to the environment. The system's ability to quickly detect and react to traffic violations also plays an important role in improving overall road discipline and reducing the risk of accidents. This demonstration aims to investigate the effectiveness of GPS-based collision avoidance, providing a better path to road safety. This innovation, which meets the need to be more efficient in accident and crime investigations, is not only based on modern technology, but also has a social and environmental purpose, being an important step towards safety and better transportation. The system integrates GPS data collection, vehicle speed monitoring, slope detection and other functions to instantly detect and react to situations. Geo-integration and Google Maps integration improve accuracy and user experience. Machine learning algorithms can identify the situation and trigger SMS alerts and alerts for quick response. Audio output and LCD screen provide immediate warning to the driver and impress the user. The system's cost efficiency, reliability and scalability allow it to adapt to different vehicles. Live monitoring provides instant feedback, while coordinated emergency response minimizes the impact of an incident. As a result, this new solution uses advanced technology to create a social, simple, efficient and effective crime prevention system and ultimately improve the safe road.

Keywords: Vehicle safety, CNN, yolo, GSM, CCTV, Blind spot detection, ADXL, ADAS.

Abstract: The project "Design and implementation of cooperative adaptive cruise control using CAN protocol" aims to enhance automotive safety and traffic flow through the implementation of an intelligent cruise control system. The system utilizes the Controller Area Network (CAN) protocol for seamless communication between vehicles and employs Microchip technology for efficient control and coordination. This cooperative approach optimizes traffic flow, reduces congestion, and enhances overall road safety by minimizing the risk of collisions.  

Keywords: CAN, CACC.

Abstract: Lakes are important ecosystems that benefit humans in many ways, including drinking water, tourism, and biodiversity. However, pollution can harm lakes, which can seriously harm human health. One strategy to lower pollution in lakes is to use automated lake cleaning boats. These boats have sensors that can locate and collect trash in the water.

Abstract: This project aims to help the medical industry.The medical industry strives to improve the delivery of key device information through the package to patients,Distributor and end users. To achieve this goal Indications For Use and user manuals have been major tools and are necessary components required in Medical Package according to Food and Drug Administration (FDA) standards.Historically there have been challenges caused by packaging information materials aspects such as manufacturing, transportation and translation. The need for extensive packaging and labelling has ultimately contributed to increased cost of manufacturing for devices. It is also important to know what information a customer needs and recognize that the safety of the consumer is of the utmost importance. The development and implementation of new technologies and procedures in a medical industry may be complicated and slow but it is a necessity to improve safety and provide maximum comfort to the end user.The existing supply chain for the pharmaceutical industry is obsolete and lacks clear visibility over the entire system. Moreover, the circulation of counterfeit medicine in the market has increased over the years. According to the WHO report, around 10.5% of the medicinal medicine in lower / middle income countries are fake and such medicine may pose serious threats to public health, sometimes leading to death.In this paper, we propose a QR Code -based model to track the movement of medicine from the industry to the users and to minimize the chances of a medicine being counterfeit.Barcodes and Two Dimensional code have been used in the medical device industry for tracking purposes; however, the focus of this thesis was using QR codes in medical device package without IFU, user guides and manuals to enhance patient safety, reduce cost and enhance the breadth of information available to the ultimate users. Access to the information was achieved by just taking a picture or scanning the QR code which was printed on a medical device package. This thesis also assesses the feasibility of implementing the QR code technology on medical device package and a case study is conducted that elaborates on the cost analysis

Keywords: QRCode, AndroidApp, Medicine, Android, Company, Dealer, Distributor, counterfeit.

Abstract: Addressing the often-overlooked issue of mental health in India, Asha emerges as a groundbreaking AI assistant with a mission far exceeding typical virtual support. Recognizing the cultural stigma surrounding mental wellbeing, Asha is designed to be a compassionate friend, guiding individuals on their path towards emotional peace and personal growth. It tackles several key aspects of mental health struggles. Firstly, Asha personalizes its interactions using advanced artificial intelligence. This allows it to tailor its responses and support to your specific needs. Asha can identify your emotional state and offer relevant coping mechanisms or activities, whether it’s calming exercises for anxiety, mood-boosting techniques for low days, or simply a listening ear for moments of overwhelm. Secondly, Asha addresses the isolating nature of mental health struggles by fostering an online community. This network allows individuals to connect with others facing similar challenges, sharing experiences and finding solace in shared struggles. By combatting feelings of isolation and loneliness, Asha’s approach supports mental well-being. Thirdly, Asha goes beyond just conversation. It incorporates interactive features and engaging exercises to keep you motivated and actively working towards your mental wellness goals. Finally, Recognizing its limitations, Asha acknowledges that professional help might be necessary. It acts as a bridge, guiding you towards finding the right therapist or mental health resources to ensure you get the most comprehensive support possible. In essence, Asha is more than just an AI assistant; it’s a multifaceted approach to mental well-being, offering personalized support, a sense of belonging through community, tools for self-improvement, and a gentle nudge towards professional help if needed. It strives to dismantle the walls of isolation and empower individuals in India to embrace their journey towards a happier and healthier life.

Keywords: Mental health in India,AI assistant for emotional support,Online community for mental well-being,Personalized mental wellness support

Keywords: Diabetic foot ulcer, YOLOv8, Machine learning, Detection

Abstract: This research presents an innovative approach to guarantee payload landing safety while establishing effective communication. Unpredictability and a lack of real-time payload transmission are two problems with traditional landing techniques. Our approach combines sophisticated sensors, smart control systems, and reliable communication protocols to overcome problems. First, to collect real-time data during descent, our method makes use of a variety of onboard sensors, such as cameras, altimeters, and inertial measurement units (IMUs). An sophisticated control system processes this data and dynamically modifies the landing trajectory according to payload specifications and environmental factors. The technology continuously optimizes the landing procedure by integrating machine learning algorithms, ensuring a safelandings even in difficult terrains. Our approach's efficacy is illustrated by simulations and experimental tests, which highlight its capacity to accomplish accurate and secure landings while preserving dependable payload communication. This method has important ramifications for a number of applications, such as cargo delivery, unmanned aerial vehicles (UAVs), and space research, where a safe landing and effective communication are critical. Key words: Inertial Measurement Units, Unmanned Aerial Vehicle, safe landing.

Abstract: This paper presents a detailed framework for an Optical Character Recognition (OCR) system employing Convolutional Neural Networks (CNNs) for recognizing optical characters. The CNN architecture demonstrates remarkable proficiency in learning various styles within input images, including handwriting and printed text. CNNs, as a subset of Deep Neural Networks, excel in recognizing and classifying specific features from images, making them widely applicable in visual image analysis tasks such as image classification, medical image analysis, and language processing. The paper outlines the essential modules and algorithms utilized in the implementation process. These modules include image processing, segmentation, feature extraction, and training/recognition. In the image processing module, steps such as grey scale conversion and image binarization are employed to prepare the input image for segmentation. Segmentation is achieved through line segmentation, word segmentation, and character segmentation, facilitating the extraction of individual characters from the document image. Feature extraction involves resizing character images and storing extracted features for further processing. Finally, the training and recognition module utilizes the Kohonen algorithm, based on Self-Organizing Maps, for training and recognizing characters. By presenting this comprehensive framework, the paper aims to contribute to the advancement of OCR systems, particularly in the context of document digitization and text recognition tasks. The proposed approach offers a robust methodology for accurately extracting and recognizing characters from various types of documents, thus facilitating automation and efficiency in document processing tasks.

Keywords: Optical Character Recognition, Convolutional Neural Networks, Image Processing, Segmentation, Feature Extraction, Kohonen Algorithm.

Abstract: The widespread use of social media has led to an alarming increase in cyberbullying incidents, causing significant psychological and emotional distress to victims. This project aims to address this pressing issue by leveraging advanced deep learning techniques, specifically Long Short-Term Memory (LSTM) networks, to detect instances of cyberbullying in social media posts. While existing research primarily focuses on established languages, there remains a notable gap in resources for emerging languages. Thus, this project seeks to bridge this gap by developing a robust model that can effectively detect cyberbullying across various linguistic contexts. The project is structured into several key phases, beginning with data collection from a popular dataset repository like Kaggle. The collected data undergoes preprocessing to remove irrelevant information and convert text into a numerical format suitable for LSTM input. Subsequently, the LSTM model is trained on the processed data and evaluated using metrics such as accuracy, precision, recall, and F1 score. The model's performance is assessed on a test set to determine its effectiveness in identifying cyberbullying instances in social media posts. Through rigorous experimentation, the LSTM model demonstrates impressive results, achieving an accuracy of 95.6% on the test set. This high level of accuracy indicates the model's efficacy in accurately detecting cyberbullying behaviour. Furthermore, the trained model can be saved and deployed to make predictions on new, unseen data, thus serving as a valuable tool in combating cyberbullying and providing support to those affected by it.

Keywords: Cyberbullying detection, Long Short-Term Memory (LSTM), Social Media, Deep learning, Text analysis.

Abstract: This research explores the application of artificial intelligence in detecting bank fraud, a problem exacerbated by the COVID-19 pandemic's shift to online operations and the proliferation of charitable funds used by criminals to deceive users. The study focuses on leveraging machine learning algorithms to analyze and identify fraudulent transactions in online banking. Its key contribution lies in developing machine learning models tailored for detecting fraudulent banking activities, along with preprocessing techniques to enhance data comparison and result selection. Additionally, the paper elaborates on methods to enhance detection accuracy, including managing highly imbalanced datasets, transforming features, and engineering new features. However, this paper proposes the use of Convolutional Neural Network (CNN) for UPI fraud detection. The CNN model is designed to analyze the spending profile of cardholders, thereby enhancing the accuracy of fraud detection. The Fraud Detection System (FDS) implemented in the bank monitors the spending patterns of cardholders, automatically blocking transactions deemed unusual and alerting the bank for further investigation. This approach minimizes the need for manual intervention and ensures swift action against fraudulent activities, thereby safeguarding users' financial assets. Keywords: Transaction, Payment, UPI, Attackers, Fraudulent, Money, Datasets, Machine learning, recognition of fraudulent operations, Convolutional Neural Networks,

Abstract: Cricket is huge in India, and the Indian Premier League (IPL) is a major cricket tournament that draws players from all over the world. Predicting IPL match outcomes is important for online traders and sponsors. We can do this by looking at various factors like the players' skills, team performance, and match conditions. In our research paper, we proposed using machine learning (ML) algorithms like SVM, Random Forest, Logistic Regression, and K-Nearest Neighbor to predict IPL match outcomes. This research shows that ML techniques can effectively predict IPL match outcomes. As cricket evolves, using advanced technologies like ML for predictions not only makes the game more exciting for fans but also helps teams and stakeholders make better decisions. Our model is a step forward in using ML to understand T Twenty cricket better.

Keywords: Indian Premier League(IPL), Linear Regression,Ridge Regression, Lasso Regression, Machine Learning, IPL Score Prediction.

Abstract: Phishing attacks pose a significant threat to cybersecurity, necessitating effective detection mechanisms. This study explores the application of machine learning algorithms for the automated identification of phishing websites. By collecting a dataset of URLs labelled as phishing or legitimate, relevant features are extracted, pre-processed, and used to train various machine learning models. The performance of these models is evaluated using metrics such as accuracy, precision, recall, and F1-score, highlighting their effectiveness in distinguishing between phishing and legitimate URLs. Continuous monitoring and updates are emphasized to adapt to evolving phishing tactics. This research provides practical insights into the application of machine learning for phishing detection, contributing to the advancement of cybersecurity measures.

Keywords: Phishing detection, Machine learning, Cybersecurity, Feature extraction, Model evaluation

Abstract: Quantum computers might be more potent than the normal classical computers and Supercomputers. Some of the specific applications like Quantum simulation, Cryptography, Optimization etc. Normal classical computers are worked based on the binary system (0,1) Whereas in the Quantum computers are worked as Quantum bit also termed as Qubit. Quantum computers use qubit, which can represent 0, 1, or any superposition of these states. This property enables quantum computers to process information in unique ways. The Qubit state can be 0&1 at the same time. When we observe that it can collapses into one of the possible states. We propose to implement the predictive capability of the Quantum Machine Learning (QML) classifier on the MNIST Handwritten Digits dataset. We deploy the model on the IBM Quantum computer using Qiskit.

Keywords: Classical Computers, Quantum Computers, Qubit, MNIST Dataset, QML.

Abstract: Skin cancer presents a significant global health challenge, necessitating early and accurate diagnosis for patient survival. However, clinical evaluation of skin lesions is hindered by long waiting times and subjective interpretations. To address these issues, deep learning techniques have been leveraged to assist dermatologists in making more precise diagnoses. In this project, we aimed to develop reliable deep learning prediction models for skin cancer classification, addressing class imbalance and facilitating model interpretation. Initially, a Convolutional Neural Network (CNN) was optimized using the HAM10000 dataset, achieving 81% accuracy with the combination of Swish activation function and RMSprop optimization. To further enhance performance, we explored advanced models such as Xception and DenseNet, anticipating an accuracy of 90% or higher. Additionally, we propose extending the project by integrating these models into a user-friendly interface using the Flask framework, enabling user testing with authentication. This comprehensive approach holds promise for improving early detection and treatment of skin cancer, ultimately reducing its morbidity and mortality. Index Terms: Skin cancer, Optimised CNN, Optimization functions, Activation functions

Abstract: This paper addresses the challenging task of brain tumor segmentation in 2D Magnetic Resonance Brain Images (MRI), recognizing the limitations of manual classification and the complexities arising from diverse tumor appearances. The comprehensive analysis employing traditional classifiers like Support Vector Machine, Multilayer Perceptron and a Convolutional Neural Network (CNN). The primary objective centers on distinguishing normal and abnormal pixels based on texture and statistical features. Notably, the CNN outperforms traditional classifiers, providing a robust foundation for accurate brain tumor segmentation. This research contributes significantly to advancing the field of medical image processing, offering a robust and efficient approach for brain tumor segmentation with room for further optimization.

Keywords: Brain tumor segmentation, Magnetic Resonance Imaging (MRI), Convolutional Neural Network (CNN), Traditional classifiers, Support Vector Machine (SVM), Multilayer perception.

Keywords: Keywords for Precision Monitoring for Parkinson’s Disease using Machine Learning include early detection, Parkinson’s disease, audio signals, Machine Learning, classifier, gradient boosting, XGBoost, hyperparameter, accuracy and precesion.

Abstract: Inside the hastily evolving instructional surroundings, colleges come upon the assignment of managing tricky examination methods. Our project centres at the development of a tailor-made organization resource planning (ERP) device for university examinations. This machine aims to streamline duties which includes scheduling, student registration, paper placing, grading, and result processing. Via careful design and implementation, our ERP system facilitates seamless coordination amongst students, college, and administrative body of workers. Key features consist of user-pleasant interfaces, automated scheduling algorithms, comfy statistics management, and real-time result generation. This paper investigates the significance of ERP systems in higher education, with a focal point on faculties. It examines the functionalities and blessings of ERP structures in streamlining administrative techniques, improving communication, and improving common performance within college campuses. Additionally, it explores challenges related to ERP implementation and integration in instructional settings, supplying techniques to conquer them. Case studies of a hit ERP implementations in faculties are highlighted, offering insights into nice practices and classes discovered. In the long run, this paper contributes to the know-how of ERP systems in higher education and gives guidance for schools considering ERP adoption or optimization.

Keywords: streamline obligations, scheduling, pupil registration, paper setting, grading, and result processing

Abstract: This work introduces a comprehensive Emotion-based playlist recommendation systems have gained significant attention in recent years due to their ability to personalize music listening experiences. In this survey, we present a novel approach where playlist management is centralized and administered through a web-based interface by an admin, while users interact with the system via a dedicated Android app. The admin is empowered to curate playlists, update content, and oversee the playlist ecosystem through the website dashboard. Concurrently, users access the system through the Android app, which offers features such as emotion detection, playlist recommendations, and seamless music playback. By leveraging emotion recognition algorithms and user preferences, our system aims to deliver tailored music playlists that resonate with users' moods and preferences, enhancing their overall listening experience. This abstract highlight the dual functionality of our system, catering to both the administrative needs of playlist management and the user-centric features of emotion-based music recommendation.

Keywords: Music player, Chatbot, Emotion, Gender, Audio, Text.

Abstract: This study assesses medical images, particularly Computed Tomography (CT) scans, for the early detection of lung cancer using processing the image, machine learning, and modern technology. The study highlights how raising patient survival rates depends on early-stage detection. Getting accurate standard performance is the primary goal. The methodology involves several processes, including dataset acquisition, data augmentation, pre-processing, selection of features, extraction of features, and CNN implementation. The outcomes of the trial indicate the precision with which our proposed technique works and how it could improve medical imaging in the existing clinical context for prevention and the therapy for lung cancer.

Keywords: Lung Cancer (LC), CT scan images, Convolutional Neural Networks (CNN)

Abstract: Since from the last few years there is a significant increase in credit card transactions are playing a vital role. Thus, it is leading to significant financial losses everywhere in present days. It is very challenging task to process the huge amount of data and it is making data sets unbalanced and complex. There are basically two major problems while handling data. It is analysed with fraud and non-fraud transactions, and it doesn’t contain relevant, appropriate, and correlated data that affects their prediction performance in a negative way. Followed by, it has involved the interest of machine learning (ML), which consists of fraud detection as a main theme. It has been involved by various ML methods such as Logistic Regression (LR), Support vector machines (SVM), Decision Trees (DT), Random Forest (RF), and K-Nearest Neighbours (KNN). However, the above methods cannot meet the excellent performance required to find and predict abnormal fraud patterns. In this project the main contribution is to provide a framework for fraud detection (FFD). Firstly, we have to overcome the unbalanced data issue, the framework uses an under sampling technique. Followed by, we have to select the relevant features by applying the feature selection (FS) mechanism. Next, Neural networks is mainly builds the ML model and it aims to handle the capability, a modified version of the Particle Swarm Optimization (PSO) algorithm, Polynomial Self Learning PSO (PSLPSO), is proposed for hyper parameters C and σ. Finally, the framework’s effectiveness is depicted in the experimental results on a transaction dataset of real credit card.

Abstract: This study explores the application of convolutional neural networks (CNNs) in biomedical image analysis for the detection of colon and lung cancer. Leveraging the power of deep learning, we aim to develop a robust and accurate system capable of identifying cancerous lesions in colonoscopy and lung CT scan images.The research begins with the collection and preprocessing of a sizable dataset containing annotated medical images. Various data augmentation and normalization techniques are applied to enhance dataset diversity and model generalization. Subsequently, a CNN architecture is carefully designed, either adapting existing architectures or crafting custom ones tailored to the unique characteristics of medical images. Training the CNN involves splitting the dataset into training, validation, and testing sets, and employing optimization algorithms to minimize a chosen loss function. Hyperparameter tuning and validation set monitoring ensure the prevention of overfitting and the optimization of model performance.Evaluation of the trained model includes rigorous testing on held-out data to assess its accuracy, precision, recall, F1-score, and AUC-ROC. Error analysis aids in understanding the model's weaknesses and identifying avenues for improvement.Ultimately, the developed model holds promise for deployment in clinical settings, pending compliance with regulatory standards. Collaboration with domain experts ensures the system's alignment with clinical needs, while continual refinement based on feedback and advancements in the field drives ongoing improvement.

Abstract: A frequent kind of arthritis, knee osteoarthritis is characterized by sclerosis, joint space narrowing, osteophyte growth, and bone deformities that can be seen on radiographs. Radiography is the most affordable and widely accessible method, and it is considered to be the best. The Kellgren and Lawrence (KL) grading technique is used to classify X-ray pictures in accordance with the progression of osteoarthritis from normal to severe. Degeneration of osteoarthritis in the knee can be slowed down by early identification, which can aid in early treatment. Regretfully, in an effort to enhance the performance of their models, the majority of currently used methods either combine or eliminate confusing grades. The objective of this research is to present an approach by leveraging an ensemble of CNN models, specifically MobileNet, ResNet, and AlexNet architectures. The choice of using a Convolutional Neural Network (CNN) for knee osteoarthritis classification is driven by its capacity to leverage deep learning techniques for medical image analysis. CNNs excel at feature extraction from medical images, making them ideal for identifying subtle patterns indicative of osteoarthritis. This approach improves the potential to automate diagnosis, reduce human error, and patient outcomes by enabling timely intervention, underscoring its relevance in the realm of medical image analysis. An Osteoarthritis Initiative (OAI) based dataset of knee joint X-ray images is chosen for this study. The dataset was split into the training, testing, and validation set with a 7.5: 1.5: 1 ratio. Our results shows that the ensemble approach significantly outperforms individual model predictions, achieving an accuracy of 96%. This improvement underscores the potential of using deep learning ensembles in medical image analysis, offering enhanced diagnostic processes in KOA classification.

Keywords: Knee Osteoarthritis (KOA), Osteoarthritis dataset, CNN, AlexNet, ResNet, Mobile Net, Ensemble model

Abstract: The Internet of Things (IoT) is a network of interconnected devices, and as this network grows, so does the need for an effective and safe protocol. Internet of Things (IoT) technologies are advancing quickly to meet the demand for the characteristics needed by applications, such as coverage area, scalability, transmission data rate, and applicability, referring to the designs of protocols. This is because of the vast range of uses and diversity of features required to meet an application. This article offers a thorough analysis of IoT protocols, including a comprehensive explanation of each protocol categorized by long- and short-distance coverage. For every set of protocols, a comparative analysis is carried out to offer insights into their traits, constraints, and behavior.

Keywords: IoT, Short Range Protocols, Long Range Protocols, LPWAN, LR-WPANs.

Abstract: At the forefront of this trend are the Internet of Things, smart cities, digital transformation of businesses, and the global digital economy. Due to the enormous amount of data collected, the strain on data storage is only going to increase, propelling the quick growth of the entire storage business. The ability to store and manage data makes cloud storage systems an essential component of the modern world. Governments, businesses, and individual users are currently actively moving their data to the cloud. An enormous volume of data can yield enormous riches. On the other hand, this raises the possibility of risks such data leakage, illegal access, revelation of sensitive information, and privacy disclosure. There are research on data security and privacy protection security, but systematic surveys on the topic in cloud storage systems are still lacking. In this study, we conduct a thorough literature review on data encryption technologies, privacy and security concerns, and relevant countermeasures for cloud storage systems. In particular, we begin by providing an overview of cloud storage, including its definition, categories, architecture, and uses. Second, we provide a thorough examination of the difficulties and specifications related to Cloud storage's data security and privacy protection systems. Thirdly, a summary of data encryption technologies and security techniques is provided. In conclusion, we address various unresolved research issues related to cloud data security storage.

Keywords: cloud storage, encryption, data security, access control, and privacy defense

Abstract: The development of LPWAN (low power wide area network) technology is gradually becoming an evolution of IoT (Internet of Things) applications, for its significant improvements of signal sensitivity and noise tolerance. In this a long-range vehicle monitoring system, based on the LoRaWAN protocol. We clarify the system parameters and determine its communication range. Finally, the communication range is concluded and a solution is proposed for setting up a Authenticated Access Control for Vehicle Ignition System and Long range Vehicle monitoring system based LoRaWAN.

Keywords: LPWAN, LoRa WAN, IOT.

Abstract: Drowsiness and intoxication are significant contributors to road accidents, posing a serious threat to public safety. This paper proposes a comprehensive system aimed at preventing fatal accidents by proactively alerting tired or emotionally distressed drivers in real-time. The system utilizes cutting-edge technologies to continuously monitor the driver's facial expressions, detecting signs of drowsiness or extreme emotional changes such as anger. Upon detection, the system takes control of the vehicle, initiates emergency measures, and alerts the driver through alarms, ensuring the safety of all occupants.

Keywords: - Chronic kidney disease, Machine learning, XgBoost classifier, Classification model.

Abstract: In today's cities, effective traffic management systems play an important role in ensuring traffic balance and reducing congestion at intersections. This article presents a new hardware-based traffic management system (STMS) designed to adjust traffic signals in response to oncoming traffic. STMS leverages advances in Internet of Things (IoT) technology to integrate sensor networks and control systems into critical interfaces to achieve real-time monitoring and signal updates. The main function of STMS is the ability to adjust the set time according to the proximity of the vehicle. When the vehicle approaches the signal, sensors detect their presence and send this information to the central control. The system then uses advanced algorithms to intelligently prioritize oncoming traffic by turning the traffic light green, thus encouraging traffic flow and reducing the potential for collisions. In fact, STMS uses a responsive approach to signal control, ensuring that the signal returns to normal operation after the vehicle has passed through the intersection. This ensures a fair distribution of green time in transit and prevents unnecessary disruption of all traffic. STMS seamlessly integrates with the existing transport system, providing the best solution for efficient signal operation and improving overall operation. The use of STMS represents a significant advance in traffic management, providing large-scale and flexible solutions for different types of urban environments. In addition to the direct benefits of easing traffic, STMS should improve road safety, reduce travel times and reduce environmental impact by reducing traffic-related vehicle emissions. As cities continue to experience urban growth and traffic congestion, innovative solutions such as STMS will play a key role in shaping the future of transportation in the city and beyond.

Keywords: Traffic management, IoT, Smart cities, Emergency vehicle prioritization, Traffic signal control, Proximity sensors, Adaptive signal timing, Urban mobility, Vehicle detection, Intersection management.

Abstract: Poultry farming is critical for global food security, yet it faces significant challenges in disease diagnosis, leading to economic losses and public health risks. Traditional methods are often time-consuming and inaccurate. Recently, deep learning (DL) techniques have emerged as powerful tools for disease diagnosis. This paper reviews the application of DL methods in poultry disease diagnosis. First, we discuss various poultry diseases, emphasizing early and accurate diagnosis. Next, we explore deep learning concepts, highlighting its ability to learn complex patterns from large datasets. We survey state-of-the-art deep learning architectures like convolutional neural networks (CNNs) optimized for poultry disease diagnosis. We address challenges such as dataset availability, model interpretability, and generalization to diverse conditions. Finally, we outline future research directions, including transfer learning and multi-modal data fusion, to enhance poultry disease diagnosis and mitigate its impact on global food security.

Keywords: Poultry, Disease diagnosis, Deep Learning, Dataset, Preprocessing, Convolutional Neural Networks (CNNs), Image classification, transfer learning, fine tuning, accuracy.

Abstract: This research paper investigates the application of deep learning models, specifically Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) networks, for predicting stock market prices. Utilizing historical data and trading volume information, our study aims to assess the performance and comparative effectiveness of LSTM and GRU architectures in capturing temporal dependencies and complex patterns inherent in financial time series data. Through comprehensive experimentation and evaluation, we analyze the predictive capabilities of both models and identify their strengths and limitations. Our findings contribute to advancing the understanding of deep learning techniques in financial forecasting, providing valuable insights for practitioners and researchers alike. By exploring the nuances between LSTM and GRU networks in stock market prediction, this study offers guidance for selecting appropriate models for future applications in financial markets.

Keywords: Stock Market, Deep Learning, LSTM, GRU, Finance.

Abstract: The project "Smart IoT-Based Fruit Chemical and Disease Detection Using Machine Learning" aims to enhance the efficiency of fruit quality assessment and disease detection through a multidimensional approach. Leveraging image-based detection and gas sensor technology, the system employs machine learning algorithms to analyse visual data and chemical emissions. The image-based detection utilizes computer vision techniques to identify visual cues associated with fruit diseases, while the gas sensor component focuses on chemical signatures emitted by fruits. By integrating these features into an IoT framework, the system provides real- time monitoring and analysis, allowing for early detection of diseases and chemical anomalies. Apart from just ensuring the accuracy of fruit quality assessment but also facilitates prompt intervention and decision-making in agricultural practices, contributing to improved crop yield and overall sustainability.

Keywords: Disease, Chemical detection, IoT, Machine learning

Abstract: The rapid proliferation of Internet of Things (IoT) devices has brought about unprecedented opportunities for connectivity and data-driven innovation across various domains. However, this surge in interconnectedness also raises significant concerns regarding security and privacy. This survey paper synthesizes a comprehensive range of literature spanning from seminal works to recent research endeavors, delving into the multifaceted landscape of IoT security and privacy. We explore various dimensions of trust management, privacy-preserving techniques, and security frameworks tailored for the IoT ecosystem. Additionally, we scrutinize the evolving threat landscape, encompassing vulnerabilities such as Heartbleed and privacy implications associated with location tracking. Drawing upon insights from diverse scholarly contributions, we aim to provide a holistic understanding of the challenges and advancements in safeguarding IoT systems against malicious exploits while preserving user privacy. By synthesizing the collective knowledge from the surveyed literature, this paper offers valuable insights for researchers, practitioners, and policymakers engaged in fortifying the security and privacy foundations of the IoT paradigm. Keywords: Internet of Things (IoT), security, privacy, trust management

Abstract: This project proposes a model to build an efficient attendance recording application that uses advanced deep learning techniques for face recognition. This project aims to reduce the time-consuming traditional attendance recording method and avoid errors such as manual errors and proxying. Although this attendance registration system may already exist, they operate in a way that uses individual student images. The program aims to use a group photo of students to identify each student and create an Excel file that professors can use to record attendance. Using advanced deep learning techniques and a native-react developed application, the application is trained to achieve maximum accuracy through a well-designed, easy-to-use graphical user interface application that would be deployed in the Play Store, allowing users to interact with it. The research involved studying and analyzing different works by different researchers, how they designed and created an application, and how we can improve it, simplifying the registration process of attendees and setting a new standard of accuracy and convenience in the attendance management system.

Keywords:  Face Detection, Deep learning technique, Deep Face, Face recognition, Android application, VGG.

Abstract: Crop diseases have grown significantly in recent years due to severe climate change and weakened crop immunity. This results in widespread crop destruction, lower cultivation, and ultimately financial loss for farmers Recognizing the illness and treating it have become significant challenges due to the diversity of diseases growing quickly and farmers' lack of expertise. The texture and visual similarity of the leaves help determine the kind of illness. Therefore, the resolution to this issue lies in the use of deep learning to computer vision. In this study, a deep learning-based model using images of both well and ill crop leaves is proposed, and it is trained on a public dataset. The model accomplishes its goal by categorizing photos of leaves into categories according to the pattern of defect.

Keywords: crop image dataset, CNN, MobileNet, ResNet.

Abstract: In today's Generation Vaccination is a crucial responsibility for parents, necessitating a systematic approach to ensure timely administration. According to NFHS-5, 2019-21, the country's full immunization coverage stands at 76.1 percent. To address this need, our project focuses on developing a vaccination website. This platform serves multiple functions Parent registration, booking hospitals in located areas, managing child vaccination schedules, keeping child records, send reminder messages through SMS. The child can be registered by the parent, booking a hospital for vaccination.

Keywords: Vaccination, Immunization, Reminder (SMS), Records.

Abstract: The healthcare industry has witnessed a rapid transformation in last few years, with the digitalization of patient health records playing a pivotal role. Electronic Health Records (EHRs) have become the standard for storing and managing patient information, offering convenience and accuracy in healthcare delivery. However, concerns regarding the integrity of EHRs persist. To tackle these issues, this project implements blockchain-based system for the private management of electronic health reports. This project will involve the enhancement of a blockchain network, a user-friendly front-end application for medical care providers and patients, and integration with EHRs.

Keywords: blockchain technology, health report, electronic health report, EHR.

Abstract: In recent years, the widespread influence of social media platforms has sparked significant concerns about their impact on mental health. This abstract delves into the intricate relationship between social media usage and mental wellness. Research indicates that while social media can foster connection and support, excessive use may lead to negative outcomes such as increased feelings of loneliness, depression, and anxiety. Moreover, the selective nature of content on social media platforms often contributes to unrealistic comparisons and self-esteem issues among users. Additionally, cyberbullying amplifies mental health risks, especially among vulnerable groups like adolescents. Despite these challenges, interventions such as digital literacy programs and promoting mindful consumption hold promise in mitigating the adverse effects of social media on mental health. Continued interdisciplinary research is crucial for understanding these complex dynamics and developing effective strategies to promote positive mental well-being in the digital age. Social media usage and its impact on mental health, specifically depression, anxiety, stress, and self-esteem, loneliness were evaluated across different platforms like Facebook, Instagram, and Twitter. The study aimed to provide a comprehensive understanding of how various types of social media usage relate to self-reported mental health indicators. Initial results suggest that individuals belonging to Generation Z perceive a more adverse effect on their mental well-being due to their use of these popular social media platforms.

Keywords: Social media platform, Mental health, Loneliness, Depression, Anxiety.

Abstract: Artificial Intelligence (AI) has rapidly become an integral part of our daily lives, transforming various aspects of society and opening up new possibilities and opportunities. However, the growth of AI also raises concerns about its impact on society and the potential consequences of its widespread adoption. This research paper provides a comprehensive overview of the impact of AI on various aspects of society, including the economy, education, healthcare, employment, and ethics. To achieve this goal, a thorough review of existing research and data on the impact of AI was conducted. The literature review focused on the economic, social, and ethical implications of AI, as well as the challenges associated with its implementation. The review found that while AI has the potential to bring significant benefits to society, it also poses challenges and risks that need to be addressed. The paper also presents a discussion of the impact of AI on different sectors of society, including healthcare, education, and employment. The analysis found that AI has the potential to improve patient outcomes and provide more efficient and effective healthcare services. It can also transform education by providing personalized and adaptive learning experiences. At the same time, the implementation of AI in the workplace raises concerns about job displacement and the potential for economic inequality. In addition, the paper evaluates the ethical implications of AI and identifies the need for responsible development and deployment of AI systems. This includes the need for ethical frameworks and guidelines to address issues such as bias, privacy, and transparency. Overall, this research paper provides a comprehensive overview of the impact of AI on various aspects of society, identifies key trends and challenges associated with its implementation, and presents possible solutions to address these issues. By doing so, this paper helps to guide policy-makers, technology experts, and the general public towards the responsible and equitable deployment of AI.

Keywords: Artificial Intelligence, Ethical Implications, Healtcare, Impact, Society

Abstract: The online gaming sector, which is rapidly expanding, heavily relies on the advancement of network technologies. The current networking technologies utilized in online gaming are undergoing a significant transformation, primarily due to the adoption of QUIC and the rise of cloud and VR gaming. At the core of online gaming lies the Internet Protocol (IP), facilitating seamless communication and data exchange between networked devices. IP plays a crucial role in delivering smooth, real-time gaming experiences, enabling players to connect, interact, and engage in multiplayer games across diverse platforms and geographical locations. Typically, the architecture of online gaming follows a client-server model, where players connect to a central server hosting the game. The gaming experience and performance are enriched by state-of-the-art technologies developed by game developers. Conducting a SWOT analysis of online gaming technology provides insights into its strengths, weaknesses, opportunities, and threats, offering a forward-looking perspective on potential future advancements. The gaming industry is anticipated to continue its growth trajectory, with the technologies driving the online gaming sector shaping the future direction of the industry in the coming decade.

Keywords: Online gaming ,Game developer , Cloud gaming , Internet Protocol.

Abstract: With the help of a machine learning (ML) model for toxic remark identification, this project presents a locally hosted social media platform that looks like Facebook or Instagram. An active online community is fostered by users' ability to create accounts, publish information, and participate in discussions. By utilizing cutting-edge machine learning algorithms, the platform can identify and eliminate harmful remarks on its own, creating a polite and secure place for users to engage. Proactive moderating is made possible via an email notification system that also instantly informs users of any offensive comments on their posts. With this study, we show how effective machine learning (ML) solutions can be at improving online safety and encouraging positive social media communication.

Keywords: Social Media, offending comment, Toxic Comment Detection, Positive social media communication

Abstract: This project presents the design and implementation of a smart blind stick prototype aimed at enhancing the mobility and safety of visually impaired individuals. The Blind Stick integrates an ultrasonic sensor, USB web camera, speakers, to provide object detection, The prototype leverages an Raspberry pi microcontroller to efficiently manage the sensor data and interactions. The ultrasonic sensor is employed to detect obstacles in the user's path, triggering a speak out to warn the user of potential collisions. The integration of a switch allows the user to initiate an emergency alert. Upon pressing the switch, Additionally, the Blind Stick prototype capitalizes on computer vision techniques through the utilization of the YOLO (You Only Look Once) framework. Connected to a PC, the Blind Stick leverages a USB web camera to capture images. Detected objects are then identified using YOLO, and corresponding audio alerts are relayed to the user through the earphones, enabling the user to understand their surroundings more comprehensively.

Keywords: Embedded system. raspberry pi, ultrasonic sensor, YOLO

Abstract: The primary causes of the significant decline in grape yield are grape diseases. Therefore, the development of an automatic grape leaf disease identification system is imperative. The remarkable results that deep learning techniques have lately obtained in a variety of computer vision challenges motivate us to apply them to the issue of identifying grape illnesses. This paper proposes an integrated method-based architecture for convolutional neural networks (CNNs). The suggested CNN architecture, or UnitedModel, is made to differentiate between healthy leaves and leaves that have common grape diseases including black rot, esca, and isariopsis leaf spot. The suggested UnitedModel can extract complementary discriminative features because it combines multiple CNNs. As a result, UnitedModel now has better representation. Using the withheld PlantVillage dataset, the UnitedModel has been assessed and contrasted with multiple cutting-edge CNN models. Based on multiple evaluation metrics, UnitedModel performs the best, according to the experimental results.

Keywords: Grape leaf disease, image processing, feature extraction.

Abstract: Grid computing systems are very large scale and can be used in internet sized environments with distributed machines across multiple organizations The resource scheduling system is the central component of the grid computing system. Resources in the grid are distributed, heterogeneous and autonomous. Scheduling in the grid environment depends on the characteristics of tasks, machines, and network connectivity. The number of jobs in the waiting line is known as the load and depending on the nature of the work it can be low, moderate, or heavy. The method of load balancing involves enhancing the computational grid systems performance so that all of the grids computing nodes are uniformly employed to the greatest extent possible to reduce execution time and increase throughput.

Keywords: Grid computing, Resource, Load Balance

Abstract: This result paper includes the electronic wheelchair that is implemented for the disabled person who cannot walk. Our system implemented eye ball controlled wheelchair is to eliminate the assistance required for the disabled person. In this system we are controlling the wheelchair by eye movements and central switch. A good resolution camera is mounted on wheelchair in front of the person, for capturing the image of eye and tracking the position of eye pupil by using any image processing techniques using Matlab platform. According to eye pupil position of disabled person, motor will be moved in required direction such as left, right, backward and forward with the help of Arduino. We are also using Ultrasonic sensor. It is mounted in front of wheelchair for safety to detect static obstacle or mobile barriers and to stop the wheelchair movement automatically. A central button switch is also mounted on wheelchair handle for emergency purpose and to stop movement in require direction if any one call to stop and someone require attention on themselves. This is independent and cost effective wheelchair system. An Arduino board is used to control whole system. Index Terms: Arduino, Computer vision library, Image Processing, Matlab, Eye gauge, Arduino.

Abstract: One of the important quality features of fruits is its appearance. Appearance not only influences their market value, the preferences, and the choice of the consumer, but also their internal quality to a certain extent. Our project presents a Computer Vision based technology for fruit quality detection. This will be implemented in python using CNN. In this project, we will use an external web cam to capture the real time image of a given fruit. This web cam will be connected to a computer device. Using our software, it will analyze the given fruit and checks whether it has any abnormalities like black or brown spots or uneven texture. These indications help us to identify the quality of the given fruit. The use of this technology can significantly improve agriculture & fruit industry as well as computer vision systems provide rapid, economic, hygienic, consistent, and objective assessment which provides people with a healthier lifestyle.

Keywords: CNN

Abstract: Computer vision makes extensive use of object detection. and crucial for variety of applications, by identifying and locating objects in images and videos. It’s a crucial technology for many applications, like self-driving cars and facial recognition. In general when objects are exposed to light then object detection can be done by using simple weights algorithm like Mobile Weight algorithm (MW) of object detection and identification and Common Objects In Context (COCO) data set. In MW algorithm the binary Images this used for comparison can’t accept larger variation of objects and it have faults in algorithm so we utilized You Only Live Once Algorithm (YOLO).It is used most frequency in variety of application by charity of image or video is required so updated Algorithm need to be used for increasing the efficiency of system so that object detection and identification in even in front view and top view of images or in video. Finally, Using the Machine learning approach we are trying to improve the accuracy of the classification tasks. More or less we are expecting 0.88 weighted average precision, 0.74 weighted average recall, 0.80 weighted average f1-score, and 90.51 percent accuracy by the proposed Machine Learning and computer vision methods.

Keywords: Computer Vision Method, Image processing, Coco data set, Mobile Weight algorithm, Classification and detection, You Only Live Once algorithm.

Abstract: In a humanity loaded of new technology, the menace of sting is constantly multiplying. In the shields assiduity, this threat was long before the use of technology. Congress legislated the shields Act of 1933 to battle the menace of sting and misrepresentation in the trade of armors. By challenging chock-full exposure, investors possess the occasion to make informed opinions previous to investing. still, Distributed Autonomous Associations( “ DAOs ”), through the use of blockchains and smart - contracts, engage in the trade of securities without completely telling the pitfalls or complying with the enrollment conditions of the Securities Act of 1933. Compliance with the burdensome conditions of enrollment , still, would destroy this new technology and system of conducting business. To avoid this reversal, Congress must amend the enrollment conditions to give an impunity for DAOs. This impunity, although reducing current enrollment burdens, must still bear DAOs to expose certain information, there by icing investors are informed previous to investing. likewise, due to the unique nature of the blockchain, smart contract, and DAOs, Congress must put a fiduciary duty on the generators of DAOs to insure compliance with the exposure conditions. Further, Congress should consider the allowance of burden - shifting following the original crowd trade. In a block- chain IoT atmosphere, when data or device authentication information is lay on a block chain, particular information may be blurted through the evidence- of- work course or declamation hunt. In this document, we refer Zero Knowledge evidence to a sharp cadence network to demonstrate that a prover without telling information similar as public key, and we've studied how to enhance obscurity of block chain for sequestration protection. Index Terms: Transaction, Transaction of insurance,bank transaction details, block-chain, Cloud -storage , Zero knowledge proof , Evolution of block chain.

Abstract: As we know India is minute most populated country and contains a broad youth populace, these days youth are warm of bikes and since of plan, they ignore wearing defensive cap. Since of these, bike mischances are extending day by day which causes passings. Major passings are due to head wounds which can be dodged by wearing a head defender. Failed and drive cases are getting to be more, which causes mischances and due to require of carelessness where an incident happens and people are gnawing the clean. These events made us make a quick defensive cap utilizing web of things which decrease the disasters and chance of passings, which has taking after highlights, the bike starts because it wereon the off chance that the rider wears a defensive capon the off chance that the rider is over failed at that point the begin will be actually off. and on the off chance that any disaster happens at that point through appear will send the alert message to the concerned watchmen. The objective of the sharp defensive cap is to supply a suggests and gadget for distinguishing and declaring disasters. Sensors, and cloud computing establishments are utilized for building the system. The disaster revelation system communicates the accelerometer values to the processor which ceaselessly screens for unconventional assortments. When an accident happens, the related focuses of intrigued are sent to the emergency contacts by utilizing a cloud based advantage. The vehicle region is gotten by making utilize of the around the world arranging system. The system ensures a tried and true and rapid transport of information relating to the incident in honest to goodness time and up dated to cloud which are gotten to by IOT.

Keywords: Statistical Description and Dispersion, Correlation, Feature Analysis, Classification, K-Nearest Neighbor, Decision Tree, Support Vector Machines, Naive Bayes

Abstract: In rural India, accessing quality healthcare can be a major challenge due to factors like remote locations, limited medical professionals, and inadequate infrastructure. To address these issues head-on, we're introducing an innovative solution an AI-assisted telemedicine robotic kiosk. Our goal is simple to revolutionize healthcare delivery by making it easy for people in rural areas to connect with expert doctors. Through advanced AI algorithms, the kiosk provides personalized consultations tailored to individual health conditions. To ensure privacy and security, we've implemented state-of-the-art authentication techniques. This ensures that only authorized individuals can access the kiosk and their health information, protecting patient privacy and confidentiality. Additionally to enhance accessibility and convenience, we've partnered with the e-sanjeevani App, a digital platform that facilitates telemedicine consultations and provides access to electronic health records. Furthermore, the kiosk enables timely medication delivery by electronically transmitting prescriptions to nearby pharmacies. By leveraging cutting-edge technology and overcoming geographical barriers, our project aims to significantly enhance healthcare accessibility and quality for rural populations in India.

Abstract: The process of soil analysis involves assessing various parameters to understand the quality and composition of the soil. This includes evaluating nutrient content, moisture levels, texture among other factors. The aim is to gather comprehensive information about the soil's characteristics, allowing for informed recommendations on treatments to improve fertility and optimize agricultural practices. To further automate and optimize agricultural practices, the system incorporates a moisture sensor. This sensor measures the moisture levels in the soil and assesses the need for water. When the system determines that additional moisture is required, it automatically triggers the release of these resources (water) into the field. To enhance this soil management system, Yolo v8 algorithm are employed. This algorithm play a crucial role in identifying and classifying unwanted plants within the crop field. By leveraging advanced image recognition and analysis, the system can distinguish between desired crops and invasive or harmful plants. Once identified, the system promptly notifies users, enabling timely intervention to address these issues and maintain the health of the crops. In addition to plant identification, the system provides real-time weather updates to users. This feature ensures that farmers and stakeholders are informed about current and upcoming weather conditions.

Keywords: Soil analysis, Yolo algorithm, Real Time Weather Updates, Moisture Sensor, Automation for Agricultural Practices

Abstract: The development of autonomous systems for drone journalism represents a significant leap forward in modern media coverage. It aims to revolutionize journalistic practices by integrating real-time object detection and tracking capabilities using the YOLOv8 algorithm. The primary objective is to create a reliable and user-friendly system that enables drones to autonomously capture footage of journalists in action, eliminating the need for manual piloting and enhancing efficiency and safety in journalistic drone operations.Utilizing the YOLOv8 algorithm, the system empowers drones to autonomously identify and track journalists, ensuring automatic footage capture from diverse perspectives and angles. Key features of the proposed system include automated flight controls, user-friendly interfaces for seamless drone operation and monitoring, and robust safety measures to minimize the risk of accidents and errors.By streamlining the technical aspects of drone piloting, journalists can focus their efforts on content creation, resulting in higher-quality and more relevant journalistic coverage. In conclusion, the integration of real-time object detection and tracking with drones using the YOLO algorithm represents a significant leap forward in autonomous drone journalism, empowering journalists to capture compelling footage efficiently and safely while enriching the storytelling experience for audiences worldwide.

Keywords: Autonomous, systems, Drone journalism, Real-time object detection, Tracking capabilities, YOLOv8 algorithm.

Abstract: This project explores the integration of Augmented Reality (AR) technology to create an innovative smart view of a college campus. Leveraging AR platforms such as ARKit or ARCore, and Unity Tool the application offers users an immersive experience by overlaying digital information onto the real-world environment. The feature of combining the real world with virtual objects enables Augmented Reality (AR) to provide a better display of information, resulting in its increasing popularity in a variety of industries. The key components include a digital campus map, markers triggering specific details and interactive elements. Creating 3D models of the campus that can be overlaid onto the real environment using AR, allowing users to explore a virtual representation of the campus. The goal is to enhance campus exploration, provide information, and foster engagement among students and visitors. Implementing QR codes at significant locations for campus views, providing users with an AR-guided experience that includes relevant information about each stop. Through careful development and user feedback, the project aims to deliver an accessible and user-friendly AR solution for an enriched college campus experience.

Keywords: Augmented reality, Campus tour, QR Code, Smart view.

Abstract: The advent of advanced healthcare software systems presents a promising avenue for revolutionizing the early detection and management of brain tumors, a critical aspect of modern healthcare. This project delves into the development of such a system, leveraging cutting-edge technologies to enhance the efficiency and effectiveness of brain tumor diagnosis and patient care. At its core, the system harnesses the power of the YOLO (V8) algorithm to enable precise detection of tumors from MRI scans, providing clinicians with invaluable insights into patient health. Moreover, the software facilitates seamless communication between patients and healthcare facilities, streamlining processes such as appointment scheduling and confirmation in real-time. Built on a robust software architecture comprising React for the frontend and Python (Flask) and .Net (6.0) for backend functionalities, the system offers an intuitive user interface that empowers users to upload MRI scans, schedule appointments, and visualize tumor detection results with ease. Integration with Firebase ensures secure user authentication, enhancing the privacy and security of patient data. By amalgamating these technologies, this project endeavors to create a user-friendly, efficient, and integrated healthcare solution that prioritizes timely diagnosis and improved patient care. The overarching goal is to address the pressing need for early detection and management of brain tumors, ultimately contributing to better health outcomes for patients worldwide.

Keywords: Brain tumor detection, MRI scan, DL, Patient engagement, Appointment scheduling, User authentication.

Abstract: The healthcare sector has witnessed a rapid digitization of patient records, leading to an exponential increase in the volume and sensitivity of healthcare data. However, ensuring the security and privacy of this data has emerged as a critical challenge due to the evolving landscape of cyber threats. To address this challenge, a novel approach that combines the scalability of cloud computing with the immutability and transparency of blockchain technology to achieve robust security for healthcare data. The proposed hybrid storage framework leverages the advantages of both cloud computing and blockchain to establish a secure and efficient data management system. In this framework, sensitive healthcare data is encrypted and stored on distributed cloud servers to ensure high availability and reliability. Additionally, a blockchain-based distributed ledger is employed to record access logs and maintain a tamper-proof audit trail of data transactions. The integration of blockchain technology enables transparent and accountable data sharing among authorized parties while preserving patient privacy and confidentiality. The results indicate that the hybrid storage model offers superior resilience against various security threats, including unauthorized access, data breaches, and tampering, thus ensuring the confidentiality, integrity, and availability of healthcare data.

Keywords: e-Government System, Blockchain, Cloud Computing, Threat Detection, Data Backup, Verification and Auditing, intrusion attacks.

Abstract: This novel technique digitizes handwritten text using Optical Character Recognition (OCR), Mobile Nets, and Convolutional Neural Nets (CNNs). The concept is to use CNNs and Mobile Nets to extract features and classify handwritten characters, with the goal of accurately understanding them. The addition of OCR technology improves the process even further by strengthening the model’s capacity to identify different handwriting styles. Combining these techniques results in a significant improvement in character recognition efficiency and accuracy, which opens up new possibilities for document digitization, language processing, and computer interaction. This paper presents a robust framework for handwritten text interpretation in a variety of applications.

Keywords: CNN, MOBILENET, AND OCR TECHNIQUE

Abstract: The pervasive emergence of deepfake technology presents unprecedented challenges to the authenticity of digital imagery, prompting the need for advanced methods in detection and mitigation. This review synthesizes insights from multiple pivotal papers, spanning diverse approaches to image forgery detection. It begins with an exploration of the intricacies and societal ramifications of deepfake technology. Navigating through methodologies like CNN-based passive tamper detection, block-based copy-move forgery detection, and ensemble approaches using advanced neural network architectures such as Inception Resnet V2, the review scrutinizes each method's distinctive strengths and limitations, providing a nuanced understanding of their efficacy against digital image manipulations. A comparative analysis reveals the variances and trade-offs inherent in these detection methodologies, offering a valuable resource for researchers and practitioners in image forensics. The abstract concludes by outlining persisting challenges in image forgery detection and suggesting prospective avenues for future research. By distilling a comprehensive overview of contemporary image forensics, this review equips stakeholders with essential insights to navigate the evolving landscape of digital image authenticity and fortify defenses against the escalating threat of deepfake manipulations.

Keywords: deepfake, image forensics, CNN, forgery detection, challenges in image forensics.

Abstract: It is said that the numerous "Night Vision" approaches are an invention that allows us to alter our vision in low light and see in complete darkness. This invention is a combination of several diverse approaches, each with unique drawbacks and focal points. The three most widely used methods are illumination, thermal imaging, and low-light imaging. There are several night vision devices (NVDs) that produce images in light levels that gradually get darker. These devices also make clear the different applications where advances in night vision technology are utilized to address various problems resulting from low light levels. Due to the increased risk associated with nighttime transportation, the device should prioritize its capacity to recognize things that pose a threat to pedestrians and animals. continue to succeed when facing the headlights of approaching cars. It has been demonstrated that the infrared system performs better than the near-infrared system. Images in the near infrared have been found to have significantly more visual clutter than images in the far infrared. It has been demonstrated that a shorter pedestrian detection distance is correlated with visual clutter. infrared images are believed to be more unusual and consequently more challenging to view, however their existence is probably due to less visual interference.

Abstract: The proliferation of deepfake videos presents a significant challenge to the integrity of digital content. To combat this threat, we propose a novel method for detecting digital deception in videos, termed "Visual Scan." Our approach integrates graph neural networks with convolutional and recurrent neural networks to effectively capture complex relationships within video frames. By leveraging a diverse dataset encompassing various deepfake techniques such as face swapping, voice synthesis, and scene manipulation, our system achieves enhanced robustness and adaptability. Moreover, we introduce a novel adversarial training mechanism to simulate real-world scenarios, enabling our model to effectively counter evolving manipulation strategies. Additionally, our system offers real-time detection capabilities, facilitating the swift identification and containment of manipulated content across online platforms. We anticipate that our approach will significantly improve accuracy levels compared to existing benchmarks in discerning between real videos and deepfakes

Keywords: Digital Deception Detection, DeepFake (DF), Recurrent Neural Network (RNN), Convolutional Neural Network (CNN), Generative Adversarial Networks (GANs), Support Vector Machine (SVM), Long Short-Term Memory (LSTM), ResNeXt.

Abstract: This paper presents the design and implementation of a secure video conferencing application aimed at facilitating seamless communication and collaboration among users while prioritizing privacy and security. The system incorporates various features including user authentication, room creation, file sharing, and encryption. Users are required to sign in or sign up to access the platform, with each user being assigned a unique token upon registration. Meeting rooms can be created by signed-in users, and only invited participants with valid tokens are granted access to these rooms. During meetings, users can share files securely through an encrypted chat box. The application employs AES with CBC algorithm for file encryption and decryption, ensuring that only authorized users within the meeting can access the shared files. PostgreSQL is utilized as the database management system, while Neon DB serves as the ORM tool for efficient database operations. Tokens act as passwords for validating authorized users, controlling access to both meetings and shared files. WebRTC technology, implemented through Agora, enables real-time video conferencing capabilities. By integrating these components, the system provides a comprehensive solution for secure and efficient virtual collaboration, safeguarding user data and communications throughout the entire process.

Keywords: Secure video conferencing, Communication, Collaboration, Privacy, Security, User authentication, Room creation, File sharing, Encryption, Token authentication, AES with CBC algorithm, PostgreSQL, Neon DB, WebRTC technology, Agora, Real-time video conferencing, Virtual collaboration, Data safeguarding.

Abstract: Technology is becoming a vital component of modern law enforcement operations, helping to maintain the integrity of police work and strengthen security protocols. Even with the advancements in digitization, security issues still exist in the systems. Sensitive data integrity and security must still be prioritized. In order to protect user data, and their complaint, and stop unwanted access, this paper presents a robust system that enables safe police complaint registration. It does this by combining strong encryption techniques with biometric verification. The proposed solution encrypts complaints submitted through online, using the Twofish encryption technique, which is well-known for its strong security features. In addition, the system includes a selfie verification with facial recognition algorithm to increase complainants’ credibility and reduce the possibility of false accusations.

Keywords: Twofish Algorithm, Data Security, Encryption Algorithm, Facial Recognition

Abstract: Hospital Management System" revolutionizes hospital management by introducing a comprehensive and automated solution for streamlined patient care. Through a user-friendly online appointment system initiated by QR code scanning, patients seamlessly engage with the healthcare process. Timely notifications keep patients informed about their appointment status, eliminating unnecessary waiting periods. Hospital Management System" revolutionizes hospital management by introducing a comprehensive and automated solution for streamlined patient care. Through a user-friendly online appointment system initiated by QR code scanning, patients seamlessly engage with the healthcare process. Timely notifications keep patients informed about their appointment status, eliminating unnecessary waiting periods. The HMS is a direct and digital interaction platform between doctors and patients. Doctors can efficiently write electronic prescriptions on-screen, which are instantly transmitted to the pharmacy for preparation. The integration with pharmacy services ensures that prescriptions are ready for collection precisely when the patient needs them. The pharmacy, in turn, sends notifications to patients, signaling the opportune moment to retrieve their medications. HMS is a digital platform that enables doctors to write electronic prescriptions on-screen, which are then sent to pharmacies for preparation. This system optimizes the healthcare journey by reducing wait times, improving prescription processing, and enhancing patient satisfaction. It also sends notifications to patients, allowing them to retrieve their medications at the right time. OpdFlow represents a transformative approach to modern healthcare management, enhancing real-time communication and coordination between medical practitioners and pharmacies

Keywords: HMS, QR code scanning, appointment status, pharmacy, OpdFlow

Abstract: Energy is one of the major issues that the world is facing in India, the supply of energy has been one of the major problems for both urban and rural households. About 60% to 70% of the energy demand of the country is met by fuel wood and agriculture residues. Solar energy is renewable source of energy, which has great potential and it is radiated by petroleum. Solar power has become source of renewable energy and solar energy applications should be un enhanced. The solar PV modules are generally employed in dusty environment which are case trophical countries like India.

Keywords: IOT, Solar String, Fault Detection, WIFI, Adafruit IO

Abstract: Globally, cardiovascular diseases (CVDs) continue to be the primary cause of morbidity and mortality. In order to enhance patient outcomes and lessen the strain on healthcare systems, early detection and intervention are essential. Changes in retinal vascular structure may have an impact on cardiovascular health, according to recent studies. Retinal pictures are a desirable source of data for predictive modelling because they provide a non-invasive way to evaluate micro vascular abnormalities. The goal of this project is to create a machine learning model that analyses retinal images and looks for patterns that could indicate heart illness. Specifically, this model uses Recurrent Neural Networks (RNNs). Because RNNs are good at processing sequential data, they can be used to better forecast the model and capture temporal dependencies in retinal pictures.

Abstract: Food waste refers to any food that is discarded, lost, spoiled, or not consumed for various reasons throughout the food supply chain, from production to consumption. So the food waste is discarded in landfills which are decomposed to produce methane, a potent greenhouse gas that contributes significantly to climate change. Wasted food represents economic losses for businesses, from the primary producers to retail establishments, impacting the overall economy. The production, transportation, and processing of food involve significant resources such as water, energy, and land. While food is wasted, there are still millions of people facing food insecurity and hunger globally. The issue of wasted food and excess food presents both a challenge and an opportunity for positive change. the proposed project leveraging OPTICS (Ordering Points to Identify Clustering Structure) for donating leftover food to needy people using machine learning to collect surplus food from major generators and redistribute it to those in need. It is a density-based clustering algorithm that helps identify dense regions in data, making it suitable for tasks where data points form clusters of varying shapes and densities. Implement strategies to divert food waste away from landfills, reducing the production of methane and other harmful greenhouse gases.

Keywords: Food waste, donation, machine learning, sustainability.

Abstract: The knowledge of the species of different fishes is of utmost importance in fisheries management, which involves monitoring fish population and their habitats to ensure sustainable fishing practices. Over the years, the knowledge regarding the species of fishes was not known widely by the people and it most depended on the human observer. But in recent years, the development of machine learning and deep learning has enabled us with a powerful tool to help ease those problems. These systems can accurately detect the fish in the images and provide us with the name of the species of fish in the image. This paper presents a comprehensive review of the fish prediction technologies and systems. This paper also introduces an approach of using deep learning and convolutional neural network(CNN) to predict the species of the fish. The proposed approach consists of three stages: data pre-processing, feature extraction and classification. The data pre-processing stage involves preparing the raw image by applying various filters and transformations. The feature extraction stage involves using pre trained CNN models to extract relevant features from the images. Finally, the classification stage involves using a support vector machine (SVM) classifier the extracted features as fish or non-fish. This paper also discusses the potential application of the proposed fish prediction methods, including fish management, environmental monitoring, and scientific research. Furthermore, this paper highlights the limitations of the fish prediction system and also discusses the future applications and scope of it. Overall, this paper gives a comprehensive overview of fish species prediction techniques and presents an approach to fish species prediction using deep learning and CNNs. The proposed method has the potential to significantly improve the fish conservation, and contribute to the sustainability of fish population and marine ecosystem.

Keywords: CNN(Convolutional Neural Network), Data Pre-Processing, Feature Extraction, SVM(Support Vector Machine)

Abstract: Currently, many countries around the world are moving towards becoming an aging society. The mental health of the elderly is one of the key challenges in an aging society. An elderly population is a special group that needs to be taken care of closely. A key area of concern for the elderly is that of mental health and many technologies can be applied in this area. One possible tool is facial expression recognition (FER) that can be used to detect emotions of the elderly for the purpose of mental health care. Emotion recognition in the field of human-computer interaction refers to that the computer has the corresponding perceptual ability to predict the emotional state of human beings in advance by observing human expressions, behaviours and emotions, so as to ensure that computers can communicate emotionally with humans. This project proposes a reminder system to help patients or old people to take medication. It also helps the users to take appointment from the needy doctor and send notification about the appointment confirmation and notify the care taker about the appointment date and time well in advance. This project recognizes the emotions of elder people using deep learning technique and send the notification to care taker so that care taker can respond to elder people very quickly. This project uses Django framework to build backend of the system and uses MySQL for persistent data storage. Android application enables a graphical user interface were end user will interact with application.

Keywords: Mental health, facial expression recognition, emotions, human expressions, detect emotions.

Abstract: The paper suggests a novel method for using hand motions to create a virtual mouse interface. There are drawbacks to the conventional mouse and keyboard interface, especially in situations when users need to engage without using their hands, like in augmented reality (AR), virtual reality (VR), and smart surroundings. It attempts to create a reliable and user-friendly system that allows people to control a computer cursor using hand movements recorded by a camera by utilizing computer vision techniques and machine learning algorithms .Cursor control, gesture recognition, and hand detection are three of the main parts of the proposed system. Finding and following the user's hand within the camera's range of vision is known as hand detection. Techniques including skin colour segmentation, backdrop subtraction, and deep learning-based object detection may be used in this procedure.

Keywords: convolutional neural networks (CNNs), virtual reality (VR), augmented reality (AR), hand gesture recognition systems.

Abstract: Legal Guide is an innovative technology-driven initiative aimed at breaking down barriers for marginalized communities in India. Anchored by a chatbot named Legal Guide, powered by Node JS version 20.1, the platform offers a multilingual legal information repository and a comprehensive legal aid directory. Utilizing React JS for seamless user interaction and SQLite for robust database management, Legal Guide ensures intuitive assistance and accessible resources for all. With a mission to promote legal empowerment and enhance access to justice, Legal Guide stands as a beacon of support and connection in the pursuit of equality and fairness for all individuals.It is a pioneering endeavor leveraging cutting-edge technology to address the systemic challenges faced by marginalized communities across India. At its core is Ollama, an AI-powered chatbot developed using Node JS version 20.1, which serves as a gateway to a wealth of legal knowledge and assistance. Backed by React JS for a dynamic and user-friendly interface, Legal Guide offers a multilingual legal information repository to bridge language barriers. With SQLite as its database backbone, the platform ensures robust data management and scalability.

Keywords: Legal Guide,Marginalized communities,Node JS,Multilingual_legal_information.

Abstract: Brain tumours are dangerous and serious disorders affected by uncontrolled cell growth in the brain. Brain tumours are one of the most challenging diseases to cure among the different ailments encountered in medical study. Early classification of brain tumours from magnetic resonance imaging (MRI) plays an important role in the diagnosis of such diseases. There are many diagnostic imaging methods used to identify tumours in the brain. MRI is commonly used for such tasks because of its unmatched image quality.  The traditional method of identifying tumours relies on physicians, which is time-consuming and prone to errors, putting the patient’s life in jeopardy. Identifying the classes of brain tumours is difficult due to the high anatomical and spatial diversity of the brain tumour’s surrounding region. An automated and precise diagnosis approach is required to treat this severe disease effectively. The relevance of artificial intelligence (AI) in the form of deep learning (DL) has revolutionized new methods of automated medical image diagnosis. As a result, good planning can protect a person's life that has a brain tumour. Using the 2D Convolutional Neural Network (CNN) technique, this project proposes Computer-Aided Diagnosis (CAD) a deep learning-based intelligent brain tumour detection framework for brain tumour type (glioma, meningioma, and pituitary) and stages (benign or malignant). CNN is used to classify tumours into pituitary, glioma, and meningioma. Then its classify the three grades of classified disease type, i.e., Grade-two, Grade-three, and Grade-four.  The performance of the CNN models is evaluated using performance metrics such as accuracy, sensitivity, precision, specificity and F1-score. From the experimental results, our proposed CNN model based on the Xception architecture using ADAM optimizer is better than the other three proposed models. The Xception model achieved accuracy, sensitivity, precision specificity, and F1-score values of 99.67%, 99.68%, 99.68%, 99.66%, and 99.68% on the MRI-large dataset. The proposed method is superior to the existing literature, indicating that it can be used to quickly and accurately classify brain tumours.

Keywords: Technology, Students, Programming language, Tamil-speaking individuals, Language barrier, Tools, Beginner-friendly, Linguistic needs.

Abstract: A miner’s health and life are vulnerable to several critical issues, which include not only the working environment, but also the after effect of it. To increase productivity and reduce the cost of mining along with consideration of the safety of workers, an innovative approach is required. To monitor the concentration level of harmful gases, semiconductor gas sensors are used. Due to many reasons miner’s falls and lose consciousness also proper treatment has not provided them at that time, so number of miners are died. To overcome this problem the system provides emergency alert to the supervisor if a person falls by any reason.

Keywords: NFT Based Ticketing System, ERC721 standard, Blockchain technology, Ticket authenticity

Abstract: Our system uses camera feedback at the intersection using image processing and artificial intelligence to calculate traffic speeds. By optimizing the control of traffic lights according to the speed of the vehicle, we aim to reduce congestion, improve traffic flow and reduce pollution. In addition, our project eliminates ambulance delays by using an ambulance tracking system at intersections. The system detects the arrival of ambulances and adjusts traffic lights to speed up their passage, shortening response time. Geocoding facilitates operational efficiency by converting location data into location information for ambulance navigation.

Abstract: Stroke is a critical and life-threatening medical condition that necessitates early detection and intervention to mitigate its impact. This project presents a stroke prediction model using the K- Nearest Neighbors (KNN) algorithm, a popular machine learning technique known for its simplicity and effectiveness in classification tasks. In KNN algorithm is applied to classify dataset into two categories.  First is at high risk of stroke and second is at low risk of stroke.  The objective of this project is to develop a reliable and accurate prediction system that can assist healthcare professionals in identifying individuals at risk of stroke. The dataset used in this project comprises various demographic, clinical, and lifestyle features of a diverse group of individuals, including age, gender, hypertension status, marital status, heart disease history, work type, smoking habits, and more. The project findings indicate that the KNN-based stroke prediction model achieves promising results in terms of accuracy, sensitivity, and specificity. This suggests that KNN can be a valuable tool for identifying individuals who may be at risk of stroke, allowing for early intervention and preventive measures to be taken.

Keywords: stroke, k-nearest neighbors, logistic  regression, random forest, machine learning algorithms

Abstract: In contemporary manufacturing, ensuring product quality is paramount. This project introduces Deep Defect Net, a novel deep learning framework designed for the automated identification of defects in manufactured products. The objective is to revolutionize quality control processes by leveraging the capabilities of deep neural networks to discern and classify defects with unprecedented accuracy and efficiency.

Keywords: convolutional neural network (CNN) architectures, Deep learning, Semiconductor,

Abstract: We propose a novel approach for language identification, specifically tailored for the challenging task of distinguishing homophonic short utterances. Homophonic utterances, where different languages produce similar sounds, pose a significant challenge in multilingual speech processing. We introduce a Convolutional Neural Network (CNN) architecture optimized for extracting discriminative features from audio segments. These homophonic utterances, characterized by similar sounds across different languages, are notoriously difficult to distinguish and thus require specialized techniques in multilingual speech processing. Experimental results demonstrate the superiority of our CNN-based approach in language identification, making it a valuable contribution to the field of multilingual speech processing. The experimental study was carried out on a real-time dataset comprising Hindi, Kannada, Telugu, Marathi, and several other languages. In addition to the CNN-based approach, we also employed three traditional classifiers: Deep Learning, Convolutional Neural Network, and others. Experimental evaluations underscore the effectiveness of the CNN-based approach, showcasing its ability to achieve impressive accuracy in identifying languages within homophonic contexts. To provide a comprehensive assessment, we implemented approaches for different duration intervals, including 5 seconds, 10 seconds, and 20 seconds. This innovative methodology addresses a pressing challenge in language identification, particularly in the context of homophonic utterances, and offers a promising solution for multilingual speech processing. Through rigorous experimentation and comparative analysis, our approach demonstrates notable advancements in accuracy and performance, thereby contributing significantly to the field of language identification and multilingual speech processing.

Keywords: Multilingual speech processing, CNN, Real-time datasets Hindi, Kannada, Telugu, Marathi, tamil ,Urdu, 5 sec, 10 sec, 20 sec, Deep learning.

Abstract: The concept of a "personal virtual doctor" refers to an innovative approach in healthcare that utilizes virtual or digital technologies to provide personalized and accessible medical guidance and support. This virtual doctor operates in a digital realm, machine learning, and advanced algorithms to interact with users in a manner like a human healthcare provider. Analyzes data to provide insights into the user's health status and potential areas for improvement. In the realm of technological innovation, the convergence of machine learning and healthcare has led to the development of a groundbreaking solution – the Personal Virtual Doctor. This project harnesses the capabilities of machine learning in Python to create an advanced system capable of predicting diseases based on user-input symptoms. Augmented by a sophisticated healthcare chatbot, the application offers an interactive platform for users to describe their symptoms and receive real time information about potential health issues. With a user-friendly interface and a commitment to privacy and security, this project signifies a transformative leap towards a more informed and proactive approach to personal health. The Personal Virtual Doctor is poised to revolutionize healthcare, empowering individuals to take charge of their well-being through the amalgamation of technology and medical expertise. Challenges faced by many people are looking online for health information regarding diseases, diagnoses, and different treatments. If a recommendation system can be made for doctors and medicine while using review mining will save a lot of time. The idea behind recommender system is to adapt to cope with the special requirements of the health domain related with users. The development and implementation of a personal virtual doctor aim to enhance healthcare accessibility, provide timely information, and empower individuals to take a more proactive role in managing their health and well-being.

Keywords: Personal Virtual Doctor, Disease prediction System, Multiple disease prediction

Keywords: Facial Recognition, Voter Verification, Electoral Fraud, Firebase.

Abstract: Kidney stones, also known as renal calculi, are solid deposits that form in the kidneys due to minerals and salts. They can range from tiny grains to larger formations and often lead to intense pain and complications. Dehydration, dietary factors, genetics, and certain medical conditions are common causes. Symptoms of kidney stones include severe back or side pain, blood in the urine, frequent urination, and nausea. Diagnosis involves looking at medical history, doing a physical exam, using imaging tests like non-contrast CT scans or ultrasound, and performing laboratory tests such as urinalysis. Treatment options depend on the characteristics of the stone and may include pain medication, increased fluid intake, medical expulsion therapy, shock wave lithotripsy, ureteroscopy, or surgery. It's important to have regular check-ups and make lifestyle changes to prevent kidney stone recurrence. Early detection and proper management are key to reducing symptoms and preventing complications. Machine learning can help in detecting kidney stones by analyzing medical imaging data, like CT scans or ultrasound images.

Abstract: Crimes are at rise and becoming difficult for police to identify and catch the criminals. This increasing crime rate can be reduced by giving alert to the person before its occurrence. Our Proposed System will use Face Recognition Algorithms to detect Criminals and will also use face expressions detection to detect expressions of the person. Face Recognition and Face Expression begins with extracting the coordinates of features such as width of mouth, width of eyes, pupil, and compare the result with the measurements stored in the database and return the closest record (facial metrics).The system will be running in detection mode [i.e scanning] .If a person is feeling uncomfortable with people surrounded by him/her, can scan their face and find out whether that particular person has any crime record or not. If the person is having a crime record then the word criminal is displayed on the screen. If the person is not having any crime record but still he/she is feeling uncomfortable then they can use the emergency button, click on the emergency button then the location of user, image of the suspect and user, and a message for rescue is sent to the volunteers of the system. Here volunteers are the persons, who will register into the system in order to help the people in need.

Keywords: Face recognition, Crime records, Machine Learning  paradigms,  Neural networks.