Artificial Intelligence (AI) is emerging as a powerful agent of change in education, reshaping traditional learning methodologies and enhancing student engagement through intelligent learning platforms. This paper explores the evolution of AI, its applications in education, and its impact, both positive and negative on the learning process. AI-driven tools such as adaptive learning systems, intelligent tutoring, automated assessments, and virtual teaching assistants have revolutionized classroom experiences, making education more personalized and data-driven. This paper includes an introduction to AI, literature reviews, working of AI, the impact of AI in education; positive and negative, benefits of AI, comparisons, methodology, and conclusion.

Read More...

This project introduces a web-based restaurant table booking system developed using PHP and MySQL. It aims to improve efficiency, reduce scheduling conflicts, and offer a better customer experience by digitizing the reservation process. Traditional booking methods such as manual registers or phone-based reservations often lead to mismanagement, overbooking, and longer customer wait times. The system allows users to select a date, time slot, and number of guests, and reserve a table through a simple and responsive web interface. The admin panel provides restaurant staff with an overview of daily bookings and management of table availability. This application utilizes HTML, CSS, and JavaScript on the client side, with PHP handling server-side operations and MySQL serving as the database, enabling efficient real-time reservation management. User authentication secures administrative functions such as managing table data or editing reservations. This lightweight, browser-accessible solution is especially beneficial for small and mid-sized restaurants. Future versions may include SMS/email confirmations, multi-branch support, and calendar API integration. The project offers an effective, scalable alternative to traditional reservation systems.

Read More...

The widespread use of mobile applications has woven them into the fabric of our daily lives, whether for personal tasks, corporate responsibilities, or just for fun. Take, for instance, the “Domestic Android Application for Home Services.” This corporate mobile app is designed for Android users and serves as a bridge between clients and service providers, utilizing the KNN algorithm to make connections. When clients request home services, the app pinpoints their location using latitude and longitude, ensuring that the nearest service provider is dispatched to meet their needs. By building on the existing “Facility Kart” application, which doesn't incorporate KNN, this new app opens up exciting possibilities, like integrating maps for easy drag-and-drop location changes and expanding its availability to other mobile operating systems beyond just Android.

Read More...

In the age of contactless human-computer interaction, Smart Sketch introduces a real-time, gesture-controlled virtual drawing application that enables users to draw, select colors, erase, and change backgrounds using only hand gestures. Built using OpenCV and Mediapipe, the system captures hand landmarks via a webcam and processes them to perform intuitive drawing actions. The application features a user-friendly GUI-based login system developed with Tkinter to ensure secure access. Drawing is controlled through the relative distance between the thumb and index finger, enabling pinch-based activation and deactivation. Color and tool selection are managed by positioning fingers within pre-defined screen zones. Additionally, swipe gestures allow background switching, enriching the user experience. The system overlays drawings on customizable backgrounds in real-time and is optimized for performance across common hardware setups. Smart Sketch demonstrates how existing computer vision technologies can be adapted to create an engaging, touchless sketching interface for educational, creative, and accessibility-focused environments.

Read More...

Urban air pollution presents a critical public health and environmental challenge, particularly in fast-developing cities such as Delhi, Bangalore, and Hyderabad in India. This paper presents a comprehensive trend analysis of the Air Quality Index (AQI) in these cities over 2017-2024, integrating meteorological parameters including temperature, precipitation, wind speed, and atmospheric pressure. Leveraging MATLAB's computational capabilities, we employ advanced analytical techniques such as time-series analyses, seasonal trend decomposition, and network bias correction via the Rolling Change Method. Our findings reveal persistent high pollution in Delhi affected by seasonal atmospheric stagnation, rising pollution trends tied to urban expansion in Bangalore, and moderate yet seasonally variable pollution in Hyderabad. Meteorological factors, especially wind speed and precipitation, emerged as significant mitigators, while atmospheric pressure and temperature inversions exacerbated pollution entrapment. This research underscores the importance of city-specific air quality management and demonstrates the utility of MATLAB-based frameworks for environmental data analysis, fostering informed urban environmental policy-making for sustainable development.

Read More...

Speech-to-text technology has become a cornerstone in many areas such as virtual assistants, automatic transcription, accessibility tools, and real-time communication solutions. Our growing reliance on voice interactions has sparked significant advancements in AI-powered speech recognition systems. However, achieving high transcription accuracy presents a real challenge, especially with external factors like background noise, varying accents, crowded environments, and subpar audio quality. The rise of deep learning and AI-enhanced audio cleaning techniques—like noise reduction, echo cancellation, and speaker diarization—has notably improved transcription accuracy. Even so, we're still not quite reaching that human-like transcription quality, particularly in noisy settings or with low-resource languages. This paper takes a deep dive into AI-driven speech-to-text (STT) systems, exploring their evolution and impact on transcription accuracy. We compare various AI-driven STT models, such as OpenAI Whisper, Mozilla DeepSpeech, IBM Watson Speech-to-Text, and Google Speech-to-Text API, analyzing how they perform in different environments. By testing these models under various acoustic conditions, we assess their capabilities against the real-world problems they might face. We evaluate transcription quality using key metrics like Word Error Rate (WER), Signal-to-Noise Ratio (SNR), and latency issues. Additionally, we look into AI-based audio preprocessing methods to see how they help reduce errors and boost transcription quality. This includes examining deep learning noise reduction techniques, adaptive filtering, and speech enhancement models to understand how audio processing before transcription can mitigate external disturbances. Moreover, this study delves into the practical uses of AI-powered STT across various sectors, including healthcare for medical transcriptions, education for lecture notes, and customer support for automated call center services.

Read More...

Fused Deposition Modelling (FDM) in Additive Manufacturing (AM), has emerged as a versatile and efficient production method, enabling the fabrication of complex geometries with minimal material waste and reduced post-processing. This study investigates the influence of two key FDM process parameters printing speed and layer thickness on the mechanical behaviour of 3D printed components made from ABS and PLA. Standard tensile test specimens were modelled as per ASTM D638 Type 1 and printed using varying parameter combinations. The objective is to analyse the effects of these settings and to compare findings with trends reported in earlier research. Previous studies have demonstrated that parameter selection plays a important role in determining the mechanical properties of 3D printed parts. For instance, ABS samples printed in the axial direction with a 0.3 mm layer thickness exhibited improved tensile strength compared to those printed in the lateral direction, where void formation was more prevalent. Other research focused on optimizing printing temperature and infill density to enhance the tensile performance of ABS parts, especially in automotive applications, using ASTM D638 standards for evaluation. Additionally, studies incorporating design of experiments approaches found that mechanical behaviour can be significantly affected by combinations of infill pattern, layer thickness, and infill density, where gyroid patterns and fine layers showed enhanced strength characteristics. Unlike broader investigations that explore multiple parameters simultaneously, this study focuses solely on printing speed and layer thickness to offer a more detailed and controlled analysis. By incorporating both ABS and PLA materials and comparing results with previously published work, the study seeks to contribute to the understanding of parameter-specific behaviour in FDM printing and support future efforts in process optimization for functional and load-bearing applications.

Read More...

In the time where data breaches are increasing, the need for advanced security systems have become more important. This project provides an enhanced machine learning-based system designed for security measures by applying a multi-layered approach to detect user personas and prevent data leakage. At its core, the system uses various Machine Learning models such as Random Forest classifier combined with deep learning techniques to analyse complex behavioural and transactional data. This integration allows for more enhanced detection of user personas, adapting dynamically to increasing security threats. Further advancements include the implementation of real-time data processing and anomaly detection algorithms that monitor data for unusual patterns, alerting potential breaches before they occur. The project also incorporates encryption protocols and ensure the integrity and confidentiality of sensitive data across various networks. The project extends beyond traditional security measures by using predictive analytics to forecast potential security lapses, enabling appropriate action. This approach ensures a robust defence mechanism against cyber threats, significantly enhancing organizational security postures.

Read More...