The Water and Electricity Tracking App has the purpose of assisting users to monitor and organize their daily use of water and electricity in a convenient way. Utilizing an organized and convenient method, the app allows users to monitor their resource consumption manually and maintain their patterns of consumption under observation throughout the long run. This approach helps enhance awareness and encourages responsible usage, thus allowing people to save. The app is easy to use, from sign-up and log-in, with safe (...) storage of user-specific data. Once logged in, the user is able to input their water and electricity consumption for the day, safely stored for use at a later time. To maintain users' adherence to desired levels of consumption, the app provides instant feedback by sending reminders and alerts whenever the user goes past thresholds of usage. These alerts prompt consumers to remain within their limit, resulting in sounder. Another central aspect of the app is its feature of data visualization, enabling users to visualize intricacies in their consumption patterns. By displaying previous consumption figures in a readable and simple manner, i.e., through charts and graphs, the app allows users to identify trends and make rational decisions to curtail unwanted consumption. Further, timely notifications and nudges serve as the mechanisms of encouragement, offering consistent usage behavior. Through the integration of manual inputting of data, safe storage, goal monitoring, visual inspection, and reminder notices, the Water and Electricity Tracking App provides the user with a means of controlling their usage of resources. Through this systematic approach not only will one save on utility bills, but it also serves to help aid in overall efforts towards sustainability in the environment, making it an essential app for anyone looking to be more responsible in use. (shrink)

Conventional perimeter-focused security approaches have become inadequate for today’s evolving wireless networks. Zero Trust Network Access (ZTNA) provides a more effective solution by implementing continuous validation of both users and devices. This model strengthens network defenses through strict identity authentication, traffic segmentation, live activity monitoring, and dynamic access rules. Such mechanisms help defend against threats like impersonation and unauthorized internal movement. ZTNA integrates multi-factor authentication and machine learning to identify unusual behavior and apply intelligent, context-specific access controls. Permissions are granted (...) based on current conditions such as user role and device health. By limiting access to only what is required, it reduces potential vulnerabilities. Its flexibility makes it suitable for both remote and hybrid work environments, offering security beyond traditional perimeters. As digital infrastructures become more complex, ZTNA plays a crucial role in building robust and adaptive wireless communication systems. With its "never trust, always verify" approach, it brings a higher level of assurance and reliability to network security. (shrink)

The Smart Crop Monitoring System integrates IoT hardware and AI-driven software to facilitate intelligent irrigation and crop disease prediction. Sensors like soil moisture, rain, temperature, and humidity feed realtime data to an ESP32 microcontroller. Based on this, motors control water and fertilizer supply. On the software side, a CNN-based leaf disease detection model trained on real datasets analyzes plant images and recommends remedies. These insights are relayed to the ESP32, which automates the response. The system supports sustainable farming by reducing (...) manual effort, minimizing resource use, and enhancing productivity. (shrink)

The research work explores advanced techniques in object detection frameworks, specifically focusing on Faster R- CNN and YOLOv8, to enhance the accuracy of military aircraft detection. Leveraging the multi-stage Faster R-CNN architecture, we investigate the impact of adjusting anchor ratios, systemati- cally evaluating their influence on the model’s ability to detect objects with diverse shapes. Departing from the conventional anchor-based approach, YOLOv8 introduces anchor-free detec- tion, eliminating predefined anchors and increasing robustness to object scale variations. Our experimentation includes anchor ratio (...) configurations and explores the innovative components of YOLOv8, such as the CSPDarknet53 backbone and the C2f module. It aims to contribute insights into optimizing military aircraft detection accuracy, catering to the dynamic and critical nature of aerial threats. By providing a comparative analysis of Faster R-CNN and YOLOv8, we shed light on their strengths and weaknesses, offering valuable considerations for deployment in military contexts. The paper’s findings contribute to the continual advancement of computer vision applications in national security, emphasizing the importance of automated detection for efficient and timely threat identification. (shrink)

This paper presents a transformer-based system for automatic detection and classification of human organs from medical images. Leveraging Vision Transformer (ViT) architecture, the proposed model captures both local and global anatomical structures from high-resolution input images. The approach is trained on publicly available datasets using transfer learning and shows significant improvements in accuracy and processing time. The system has potential applications in diagnostics, surgical planning, and medical education, offering real-time assistance and reducing human error..

This paper presents a transformer-based system for automatic detection and classification of human organs from medical images. Leveraging Vision Transformer (ViT) architecture, the proposed model captures both local and global anatomical structures from high-resolution input images. The approach is trained on publicly available datasets using transfer learning and shows significant improvements in accuracy and processing time. The system has potential applications in diagnostics, surgical planning, and medical education, offering real-time assistance and reducing human error..

Due to the increase in harassment and assault, women's safety is still a priority. The manual SOS activation employed by most safety apps may not be feasible in emergency situations. To bridge this gap, we developed the ingenious Android app Saviour using Kotlin in Android Studio.or safe user logon and real-time data storage, the application makes use of Firebase Authentication and Firestore. It enables users to send real-time SOS messages via SMS and WhatsApp, share live location via Google's Fused Location (...) Provider, and include emergency contacts. In case of emergencies, users can also record brief videos, which are uploaded directly to Cloudinary for safe storage and simple sharing. When it is needed the most, the app sees to it that trusted contacts receive location and essential evidence. (shrink)

Post-processing techniques are frequently used extensively in traditional biomedical signal analysis approaches in order to extract significant events. However, post-processing latency can be harmful in time-sensitive applications like seizure detection, heart arrhythmia prediction, or fall detection in elderly patients. A real-time biological event detection system that does not require post-processing is presented in this study, allowing for instantaneous event identification. Through the direct integration of machine learning techniques, edge computing, and lightweight signal filtering into data collecting pipelines, we show that (...) real-time, precise event detection is possible. We demonstrate that the suggested model achieves competitive accuracy with much lower latency through our experiments on publicly available datasets, including PhysioNet and MIT-BIH Arrhythmia. (shrink)

This paper introduces a comprehensive AI-based traffic management system designed to dynamically optimize traffic light timings at urban intersections through the integration of real-time vehicle detection and adaptive optimization techniques. Employing the YOLOv4 Tiny model for precise vehicle detection from video inputs, the system offers an innovative approach to address the complexities of urban traffic flow. Coupled with a genetic algorithm tailored for traffic optimization, it adjusts green light durations based on current traffic density, significantly reducing congestion, lowering environmental impact, (...) and enhancing road safety. The study presents an in-depth analysis of the system's implementation, discussing the frontend user interface developed with React for handling video uploads and displaying results, and the backend processing managed by Flask for data orchestration and optimization. Empirical results from a deployment across several intersections in an urban setting have demonstrated an average reduction of 25% in wait times during peak hours, a 12% decrease in CO2 emissions, and a 15% drop in minor traffic accidents. These outcomes underscore the system's efficacy in real-world scenarios, providing a scalable solution for traffic management that aligns with smart city objectives. (shrink)

Agriculture is a critical sector within the economy, yet farmers encounter various challenges in accessing markets directly, depending on intermediaries that eat into their profit margins. In order to empower farmers by giving them a platform to directly contact buyers, negotiate prices, and streamline the supply chain, the report recommends a mobile app for direct market access. The application's lean architecture, which is tailored for mobile devices, guarantees strong performance in situations with limited resources. Farmers will have more market transparency (...) and control thanks to the suggested system's real-time listing, price negotiation, logistics tracking, and data analysis features. (shrink)

The Indian healthcare landscape has seen a significant resurgence in traditional systems of medicine, collectively known as AYUSH – Ayurveda, Yoga and Naturopathy, Unani, Siddha, and Homeopathy. With increasing global interest in alternative and holistic health practices, there is a growing need for efficient digital solutions that facilitate the operational requirements of startups within this sector. This paper presents the design and implementation of the AYUSH Startup Portal, a full-stack web application aimed at simplifying the registration and management processes for (...) entrepreneurs and innovators in the AYUSH domain. Developed using modern JavaScript frameworks like React.js and Node.js, with MongoDB as the backend database, this portal provides a centralized platform for startup registration, document submission, verification, and application tracking. The portal reduces administrative bottlenecks and increases transparency, contributing to a more inclusive and innovation-friendly environment for traditional healthcare ventures. The platform stands as a technological bridge between government regulatory bodies and startup ecosystems. It is designed to scale with increased usage and can integrate advanced features such as analytics dashboards, multilingual access, and investor matchmaking tools in the future. (shrink)

Dermatological diseases constitute a significant portion of global health concerns, with millions of people affected by conditions ranging from acne to life-threatening skin cancers such as melanoma. Despite the high prevalence of these disorders, access to dermatological care remains limited, especially in rural and under-resourced regions. The increasing burden on healthcare systems necessitates the adoption of efficient diagnostic support tools. This paper presents the development of an AI-based diagnostic system for the preliminary identification of dermatological manifestations using deep learning. The (...) proposed system leverages convolutional neural networks (CNNs) to classify dermatoscopic images into multiple skin disease categories. The model, trained on the HAM10000 dataset, incorporates visual interpretability tools like Grad-CAM to ensure that healthcare workers can understand the rationale behind predictions. Our results show that the model can achieve an accuracy of over 91%, with high precision and recall, making it suitable for deployment in real-world settings where immediate dermatological consultation is unavailable. The tool is deployed through a web interface and designed with privacy and ethical considerations to ensure safe, explainable, and equitable use. (shrink)

Neuromorphic engineering is an emerging multidisciplinary field that focuses on designing computational systems inspired by the human brain’s structure and functionality. Unlike traditional systems based on the von Neumann architecture, neuromorphic platforms leveragespiking neural networks (SNNs) and custom hardware like memristors to achieve parallel, event- driven processing. These systems demonstrate remarkable adaptability, energy efficiency, and real-time responsiveness. Their capabilities make them ideal for edge computing, robotics, sensory processing, and brainmachine interface applications. By processing input streams as they arrive, neuromorphic systems (...) support rapid and efficient decision-making similar to biological cognition. (shrink)

Many emerging applications require a higher level of flexibility, modularity, and efficiency but are dependent on advancements in communication infrastructure and distributed computing. Time-sensitive networking (TSN) standards aim at providing vendor agnostic, reliable, and deterministic communications over the Ethernet, but lack in flexibility and modularity provisions. In this context wireless communication systems are preferred given the obvious benefits in terms of increased flexibility, reduced deployment & maintenance costs, and inherent mobility support. However, the stochastic nature of the wireless medium poses (...) several challenges in achieving these benefits. In this paper we comprehensively analyze the recent standardization efforts and developments in IEEE 802.11 and 5G to enable low-latency, deterministic communications and present the current status of their integration with wired TSN. Then, we present a set of use cases that may be enabled by wireless TSN including industrial automation,automotive,or audiovisual applications . (shrink)

This work introduces the design and development of Smart Parental Surveillance, a cybersecurityoriented solution that utilizes state-of-the-art keylogging methodologies to track and safeguard children online. The system offers real-time keystroke logging, application usage, and web browsing activity to help parents detect potential threats such as cyberbullying, grooming, and access to offensive content. Designed with privacy-sensitive mechanisms and safe storage protocols, the tool promotes minimal invasiveness while assuring high oversight. In contrast to conventional monitoring tools, the system proposed combines smart filtering (...) and alerting mechanisms to differentiate normal from suspicious activities. Performance reviews confirm the efficacy of the tool in identifying initial indicators of dangerous online interactions, rendering it an important resource for child online protection programs. (shrink)

This research offers a fingerprint analysis method for non-invasive blood group testing that combines machine learning and image processing to provide precise, quick results. It eliminates the need for conventional blood testing and is portable, easy to use, and perfect for emergencies or isolated locations. In addition to increasing emergency medical responses and potential future diagnostic solutions, the system increases patient comfort and reduces dangers such as contamination.

The development of deepfake technology Shows a serious threat to the integrity of voice communications. with potential misuse in identity theft, misinformation, and fraud. The main goal of the project is to create a powerful detection system that can distinguish real sound using the audio signal to increase the security of audio files. To achieve this, we propose a novel approach that integrates Convolutional Neural Networks (CNN) and Bidirectional Long Short-Term Memory (BiLSTM) networks. The CNN component extracts spatial features from (...) audio spectrograms, identifying unique patterns that differentiate real and fake audio. These extracted features are then processed by the Bi-LSTM, which captures temporal dependencies and context, further improving detection accuracy. The proposed model was tested on the “Lj speech dataset” and “wavefake dataset” , demonstrating superior performance over conventional methods with improved generalization across various spoofing techniques. This approach is particularly applicable in enhancing security in voice authentication systems, digital forensics, and other domains requiring reliable detection of audio manipulations. (shrink)

Vision- AID is an innovative AI- powered image captioning system developed to empower visually bloodied individualities by transubstantiating visual data into meaningful audio descriptions. using the power of deep literacy, the system uses apre-trained InceptionV3 convolutional neural network to prize high- position visual features from input images. These features are also fed into an LSTM- grounded decoder model trained on the Flickr8k dataset to induce grammatically coherent and contextually accurate captions. To insure availability, the generated captions are vocalized using the (...) pyttsx3 textbook- to- speech machine, allowing druggies to hear descriptions of images in real time. The result runs efficiently on original systems and is erected using Python with TensorFlow and Keras for the modeling factors. VisionAID islands the gap between vision and understanding, making digital content and surroundings more accessible and inclusive for the visually disabled community. (shrink)

This study describes the creation of an AI-based travel planning system that implements multi-agent architecture to fully automate and customize a trip itinerary. Incorporating the Groq LLaMA language model for NLU and NLG, as well as real-time web search capabilities through SerpAPI, the system can issue evolving and up-to-theminute recommendations. Streamlit was used to build the application, which enables users to specify a travel destination, trip duration, budget, traveling style, and whether they would like to hire a local guide. Using (...) these criteria, AI agents work together to provide a comprehensive itinerary complete with accommodation details, daily activities, meals, safety tips, and other relevant information on expenses. Having access to the local guides database improves the practicality of the solution. This project highlights the advances offered by multi-agent AI systems for real-life decision support problems and demonstrates how this technology can bolster intelligent travel planning systems designed for the relative ease of the customer. (shrink)

Mental health issues among children and adolescents have become a global public health concern, exacerbated by factors like academic pressure, social comparison, and the post-pandemic impact. This project aims to develop a software-based mental health surveillance platform to assess and track the psychological well-being of students, offering early detection and personalized intervention recommendations. Built using Python, HTML, CSS, JavaScript, and Flask, the system leverages AI and machine learning, including XGBoost for classification and KMeans Clustering for grouping students based on mental (...) health data such as academic performance, attendance, behavior, and mood. The platform predicts risk levels (“Low Risk,” “Moderate Risk,” “High Risk”) and suggests timely interventions for counselors, educators, and parents. The system is accessible via a secure web portal and does not require high-end infrastructure, making it suitable for schools with varying economic backgrounds. Data privacy and ethical AI standards are prioritized to ensure secure, anonymized data storage. The ultimate goal is to create a scalable solution for early intervention, contributing to emotionally resilient youth in line with India’s National Education Policy (NEP) and digital health initiatives. (shrink)