Volume 8 Issue 5
Author's Name: Anshu Anand Jethi, Ajay Rana
Abstract— “Sentiment Analysis” is the process of extracting other people's (speaker or writer) opinions from a given original source (text) utilizing natural language processing (NLP), linguistics computing, & data mining. In sentiment analysis, sentiment classification of various parameters such as comments, reviews and products has become a significant application. For the interpretation of meaning of each and every comment, “text mining approach” is used. For understanding the meaningfulness of any content, it is important to classify them into positive and negative comments on the basis of user opinion. In the present study, researcher has performed sentiment analysis on YouTube comments on the most popular topics nowadays by using Classifier techniques.
Author's Name: J.R. Arun Kumar, Devanshi Arora, Aarti Beniwal, Siya Saini
Abstract— The pervasive use of mobile devices has led to a significant increase in the volume of SMS messages, including a substantial portion of unsolicited commercial messages, commonly known as spam. These spam messages not only clutter inboxes but also pose potential security risks. To address this issue, this research presents a comprehensive SMS spam classification model that effectively distinguishes between legitimate and spam messages. The proposed model integrates powerful machine learning algorithms and advanced natural language processing techniques to analyse the textual content of SMS messages. By carefully examining the linguistic features, such as word frequency, sentiment analysis, and the presence of specific keywords, the model can accurately categorize messages into their respective classes. The implementation of this model offers several advantages. Firstly, it significantly reduces the number of spam messages reaching users' inboxes, improving their overall mobile experience. Secondly, it helps mitigate potential security threats associated with spam, such as phishing attacks and malware distribution. By effectively filtering out spam, the model contributes to a safer digital environment for mobile users
Author's Name: Mohit Sharma, Rukhsar Khan, Komal Arora, Ritika Beniwal, Sneha Sharma, Megha Saini
Abstract— The Rental Property Management System is a comprehensive web-based platform designed to streamline the management and interaction with rental property listings. This project enables users to perform CRUD (Create, Read, Update, Delete) operations on property listings, providing a seamless interface for managing rental properties. Each listing includes an integrated review section, where users can share their feedback and experiences. To enhance the user experience, the platform incorporates map integration for easy location visualization, allowing users to view property locations in a geographical context. Additionally, leveraging Natural Language Processing (NLP), the system performs sentiment analysis on user reviews. This feature provides a detailed sentiment summary, showcasing the percentage of positive, negative, and neutral comments, offering valuable insights into user feedback and property performance. The system’s intuitive interface, robust functionalities, and intelligent analysis make it a valuable tool for property managers and tenants alike, fostering transparency and informed decision- making in the rental property market.
Author's Name: Sisira S Nair
Abstract— Today the requirements for minimizing the delay, area, and power of adder circuit improve the efficiency of whole system which drives the technology to the next level. Even though the Ca Select Adder (CSLA) occupies more area used instead of ripple carry adder to avoid propagation delay. In other models Binary to Excess Converter (BEC) based Carry Select Adder was also used which uses less number of logic resources than conventional CSLA. But these CSLAs are not more efficient because it rejects one sum after the calculation. So the delay was not more effectively reduced. In order to overcome this problem the reduced logic CSLA is used. But by using Gate Diffusion Input (GDI) Technique can give less delay than this recently proposed reduced logic CSLA. The proposed technique provides low power consumption, less propagation delay. By using this GDI based CSLA the number of transistors required for the circuit also minimized. So an efficient adder design can be achieved through this technique.
Author's Name: R.Anusuya, Rahul Gupta, Sourabh Meena, Narayan Somvanshi
Abstract— In the digital age, understanding public sentiment is crucial for businesses, researchers, and policymakers. Sentiment analysis, or opinion mining, is a field of natural language processing (NLP) that identifies and extracts subjective information from text. This project aims to develop a sentiment analysis system to classify user comments as positive or negative. While trained using the IMDB movie reviews dataset, the system is designed to handle various types of user comments across different domains. The project involves data collection, preprocessing, feature extraction, model training, and deployment. The IMDB dataset, containing 50,000 labeled movie reviews, is used for training. The core of the project is the model training phase, where a machine learning model is trained to classify the sentiment of the text. An LSTM-based model is chosen for its ability to handle long-range dependencies and context in text, providing superior performance in sentiment analysis tasks. The trained model is then deployed to a production environment using a Django web application, allowing real-time analysis of new comments or reviews.Preprocessing includes tokenization, stop word removal, and label encoding. Feature extraction converts text data into numerical features using techniques like TF-IDF and tokenization. An LSTM-based model is trained to classify sentiment, chosen for its ability to handle long-range dependencies and context in text. The trained model is deployed using a Django web application, allowing real-time analysis of new comments or reviews. This system provides a valuable tool for understanding and responding to public sentiment effectively.
Author's Name: Kaviyapal, Sudesh Kumar
Abstract— In this Paperwe designed to give a output on the display of water level in a tank and control a water pump motor as required. The reading given in the scale of 16*2 display. A priority encoder is interfaced to a decoder to get the display of water level on 16x2 display. It is built around priority encoder, BCD-to- decoder, display and a few discrete components. Due to high input impedance, priority encoder senses water in the container from its nine input terminals. In this Arduino based automatic water level indicator and controller paper we are going to measure the water level by using ultrasonic sensors. Basic needs of using ultrasonic sensor to distance measurement is based on ECHO. When sound waves are transmitted in environment or in the water tank surface then they return back to the origin as ECHO after striking on any obstacle or (water) in the tank. So we have to only calculate its traveling time of both sounds means outgoing time and returning time to origin after striking on any obstacle. And after some calculation we can get a result that is the distance in the water tank thats the way we use in our water controller system. where the water motor pump is automatically turned on when water level in the tank becomes low means empty.KEYS: Relay, Ultrasonic sensor, Arduino, Bridge Rectifier, 16 X 2 LCD, water pump,Capacitor, Resistor,Transistor
Author's Name: J.R.Arun Kumar, Kunal, Naman Gupta, Drishti Kirodiwal, Tanu Jain
Abstract— This project, titled "Farm Advisor", aims to support farmers and agricultural planners by recommending the top three most suitable crops to grow based on environmental and soil-related inputs. Additionally, the system predicts the expected yield of a selected crop based on the type of crop and the size of the farming area. Another important feature of this project is the visualization of how temperature affects crop yield, which is shown using a simple and easy-to-understand graph. The system uses machine learning algorithms to make accurate crop recommendations and yield predictions. It is developed using Python, and its data visualization libraries are used to display the results in an interactive and meaningful way. By combining crop recommendation, yield prediction, and temperature analysis, this tool helps in making better farming decisions and improving agricultural productivity.
Author's Name: Kamal raj, Abhishek Singh
Abstract—This research aims to develop analytical study of post tensioned slab with floating column. In present scenario buildings with floating columns are of typical feature within the fashionable multi storey construction practices in urban India. Such sorts of constructions are highly undesirable in building inbuilt seismically active areas. For this buildings are given floating columns at one or more storey. These floating columns are highly disadvantageous during a building inbuilt seismically active areas. The earthquake forces that are developed at different floor levels during a building got to be carried down along the peak to the bottom by the shortest path. Deviation or discontinuity during this load transfer path leads to poor performance of the building. In this paper, the critical position of floating column in vertically irregular buildings has been discussed for G+11 buildings for zone II. Also the effect of size of beams and columns carrying the load of floating column has been assessed with tendons. The response of building like storey drift, storey displacement and storey shear has been wont to evaluate the results obtained using ETABS software.
Author's Name: Arvind Sharma, Shubham Singodiya, Priyanshu Saxena
Abstract— In today’s digital age, the abundance of online learning resources often overwhelms students and job seekers, making it challenging to identify reliable, relevant, and high-quality materials for career preparation. Bodhika addresses this issue by offering a centralized platform that curates and organizes trusted resources tailored to academic and professional growth. The platform provides structured study roadmaps, job preparation websites, curated YouTube playlists, online courses, and interview preparation materials, ensuring a streamlined and efficient learning experience. Built using a modern tech stack— Next.js with Tailwindcss for the frontend, Node.js for the backend, and Mongodb for data storage—Bodhika emphasises simplicity, accessibility, and user-friendliness. Authentication is handled securely via OAuth (Google, GitHub), and the platform is deployed on Vercel for scalability. By integrating curated content with a clean interface, Bodhika empowers college students, beginners, and self-learners to focus on their goals without navigating unnecessary complexity. Future enhancements, including a progress tracker with interactive dashboards, aim to provide personalised insights and motivate users. Bodhika demonstrates how a centralised, curated approach to resource aggregation can enhance career preparation, foster self-directed learning, and bridge the gap between aspiration and achievement.
Author's Name: Rajesh Kumar, Pushpendra Chawla, Mayank Ghatawal, Aditya Sharma, Bharat Singh Balaniya, Hritik Saini
Abstract—CineMatch is a smart movie recommendation system that personalizes suggestions using the K-Nearest Neighbors (KNN) algorithm. It analyzes movie metadata like genres, cast, and user behavior to find similar movies. The system uses a React.js frontend, Flask backend, and MongoDB for data storage. Users get secure logins, personalized dashboards, and feedback options. Styled with Tailwind CSS, it ensures a responsive and sleek UI. KNN offers accurate recommendations through similarity scoring. Future plans include hybrid models, mobile apps, and social features. CineMatch is a scalable solution for smart movie discovery.
Author's Name: viraswamy, Kavana Selvi, Kavya Malar
Abstract— Online health communities keep on offering enormous assortment of clinical data helpful for clinical professionals, framework executives and patients the same. In this work we gather ongoing health posts from reputed sites, where patients express their perspectives, remembering their encounters and symptoms for drugs utilised by them. We propose to perform Summarization of client posts per medication, and come out with helpful resolutions for clinical club just as patient network initially. Further, we propose to arrange the clients dependent on their 'enthusiastic perspective'. Additionally, we will perform information revelation from client posts, whereby helpful 'designs' about the triad 'drugs-symptomsmedicine’ is done by Association Rule Mining.
Author's Name: Arjun Ramu, Mohamed Ajmal, Nidhunula
Abstract— Handwritten letter recognition using computer vision and Machine Learning Technologies has been a well pondered upon topic, since the emergence of Machine Learning, the field has undergone tremendous development. As a subfield of Computer Vision, handwritten text recognition plays an important role in helping the machines to gain humanlike characteristics. There has been many experimentations on various literatures with techniques like Multiscale partial differential operators, Support vector networks, Deep belief networks and Artificial neural networks yielding acceptable accuracies. As convolutional neural network has a strong architecture, inspired by the same architecture and selectively trying to overcome the drawbacks of CNN, we use a new robust and dynamic technology of Capsule neural networks, which yields a considerably high quality results in terms of accuracies and other metrics such as precision and specificity when evaluated.
Author's Name: Mohit Sharma, Prince, Anuj, Aditya Ranjan, Apurva Singh, Dheeraj Yadav
Abstract—In today’s digital landscape, the ability to automatically extract product-related information from diverse sources such as images, packaging, advertisements, and textual content is critical for numerous applications, including e-commerce, inventory management, and market analysis. This project aims to develop a hybrid system that leverages Computer Vision (CV) and Natural Language Processing (NLP) to identify and extract product entities from both visual and textual data. The integration of CV and NLP allows for contextual understanding and improves the accuracy of entity recognition, even in complex and noisy data environments. The outcome is a scalable, intelligent framework that can automate product cataloging, support recommendation engines, and enhance user search experiences.
Author's Name:Bhaingu Singh, Pooja Sarada
Abstract— The world is experiencing an evolution of Smart Cities. These emerge from innovations in information technology that, while they create new economic and social opportunities, pose challenges to our security and expectations of privacy. Humans are already interconnected via smart phones and gadgets. Smart energy meters, security devices and smart appliances are being used in many cities. Homes, cars, public venues and other social systems are now on their path to the full connectivity known as the “Internet of Things.” Standards are evolving for all of these potentially connected systems. They will lead to unprecedented improvements in the quality of life. To benefit from them, city infrastructures and services are changing with new interconnected systems for monitoring, control and automation. Intelligent transportation, public and private, will access a web of interconnected data from GPS location to weather and traffic updates. Integrated systems will aid public safety, emergency responders and in disaster recovery. We examine two important and entangled challenges: security and privacy. Security includes illegal access to information and attacks causing physical disruptions in service availability. As digital citizens are more and more instrumented with data available about their location and activities, privacy seems to disappear. Privacy protecting systems that gather data and trigger emergency response when needed are technological challenges that go hand-in-hand with the continuous security challenges. Their implementation is essential for a Smart City in which we would wish to live. We also present a model representing the interactions between person, servers and things. Those are the major element in the Smart City and their interactions are what we need to protect.
Author's Name: Arvind Sharma, Preeti Jain, Jai Khurana, Abhay Patel
Abstract—Online voting system leveraging blockchain technology to ensure secure, transparent, and tamper-proof elections. Traditional voting systems face numerous challenges such as voter fraud, manipulation, and lack of transparency. Blockchain technology, with its decentralized and immutable nature, offers a robust solution to these issues. The proposed system integrates a user-friendly web interface for voters to cast their ballots and a blockchain backend to record votes in an immutable ledger. Each vote is encrypted and stored in a block, ensuring that it cannot be altered or deleted. The decentralized nature of blockchain distributes the data across multiple nodes, preventing single-point failures and unauthorized access. The system ensures voter anonymity while maintaining a transparent and verifiable election process. Smart contracts are employed to automate the election process, from voter registration to vote counting, ensuring efficiency and reliability. This project demonstrates the feasibility of blockchain in modernizing the voting process, enhancing security, and restoring public trust in electoral systems. The implementation includes comprehensive testing and evaluation to validate the system's integrity, scalability, and performance. This comprehensive approach aims to address existing electoral issues and provide a future-proof solution for secure voting.
Author's Name: Sitarama, zarika Begam ,Pranav Kumar
Abstract— Stress is a state of mental pressure for particular person facing troubles from environmental and social well-being which leads to so several diseases. Student age is the significant period because at this time youth faces heaps of changes in his/her life. They are probable to be the elites in the society. Thus, they should improve their stress management abilities so as to live a fit life after entering the society. When a youngster enters into the youth age, they require to not only adapt themselves to the fresh life and new surroundings but also be familiar with many new people, events, and things. The life pressure on them is considerable. Consequently, understanding the sources of stress between them and how they can cope with the stress is very significant. The researcher found that the pressure mainly comes from academic tests, interpersonal relations, relationship problems, life changes, and career exploration. Such pressure may usually cause emotional, physical, and behavioral troubles. This study finds the causes of stress among school students. So, after identifying causes the researcher suggests that extra emphasis can be given to development stage of child into adolescence. They should be brought up in the encouraging environment. More emphasis should be given to the outdoor activities and create hostile learning environment by minimize the harmful impact of stressors. The findings will help the individual students, scholars, lecturers, career and counseling centers.
Author's Name:J.R. Arun Kumar, Bhanu Gupta, Nitin Kumar, Deepanshu Choudhary, Lakshay Jain
Abstract— The Job and Internship Recommendation System is an intelligent platform leveraging machine learning and modern web technologies to enhance job discovery. By analyzing user profiles and resumes through NLP, the system generates personalized recommendations. Built using React.js and Node.js/Flask, with MongoDB and IPFS-like resume storage, it includes employer dashboards, secure authentication, and real-time tracking. This paper presents the system architecture, implementation details, and performance evaluation, showcasing its ability to transform recruitment processes with personalization and intelligent automation.
Author's Name: Kalyani Rajan, Barath Kumar , Dinesh Kumar, Ramaraj, Kadiresan
Abstract— An Artificial Neural Network (ANN) is an information processing system that is inspired by the biological nervous systems, such as the brain, process information. The key element of this system is the structure of the information processing system. It is composed of a large number of highly interconnected processing neurons working in unison to solve specific problems. ANNs, like human, learns by example. ANN is configured for a specific application, suchas pattern recognition or data classification, through a learning method. Learning in biological systems involves adjustments to the synaptic connections that exist between the neurons. This is true of ANNs as well. This paper gives overview of Artificial Neural Network, working & learning method of ANN. It also explains the application and advantages of ANN.
Author's Name: Arvind Sharma, Sanjay Saini, Sanjay Kumar, Shahnawaz Hussain, Pankaj Joshi, Shubham Saini
Abstract— This hospital management website integrates advanced AI technologies to streamline healthcare operations, enhance patient care, and optimize resource allocation. The platform offers a unified interface for managing patient records, appointments, staff scheduling, and inventory, while AI-driven tools provide predictive analytics for patient admission forecasting, automated diagnostics support, and personalized treatment recommendations. Natural Language Processing (NLP) enables efficient clinical note transcription and patient interaction analysis, while machine learning algorithms improve decisionmaking through real-time data insights. Features like AI-powered chatbots for patient triage, automated billing error detection, and real-time monitoring of critical cases reduce administrative burdens and elevate care quality. Designed with security and compliance in mind, the system ensures HIPAA/GDPR adherence. By merging operational efficiency with cutting-edge AI, this platform empowers healthcare providers to deliver faster, smarter, and more patient-centric services, revolutionizing modern hospital management.
Author's Name: Zuhika sameen, Manav Singha
Abstract— Respond to the increasing development of social content, deciphering the true from the fake becomes difficult. This therefore highlights the difficulty of blatant propaganda.This work examines the past and existing strategies to recognizing false information and how and why false news first appears in text formats.The whole paper will discuss reaches to language review and analysis methods and gives a three-part process for identifying false social media messages using the Naïve Bayes Classification, support vector machine ( svm )and Semantic Analysis systems.
Author's Name: J.R.Arun Kumar, Rohan Sharma, Mohan Singh, Sahil Gupta, Anmol Bhardwaj
Abstract— In today's fast-paced world, managing and processing large volumes of information is becoming increasingly challenging. This report introduces a Retrieval-Augmented Generation (RAG)-based chatbot designed to address the inefficiencies of traditional document analysis. The chatbot integrates cutting-edge natural language processing (NLP) techniques to extract, analyze, and summarize content from PDF documents. It empowers users by offering precise answers to their queries in natural language, leveraging a hybrid approach that combines dense vector search for retrieval and generative AI for response generation. The system begins by preprocessing the uploaded PDF to extract textual content and applies Optical Character Recognition (OCR) for scanned files to ensure no data is missed. The extracted data is then segmented, indexed, and transformed into vector embeddings using pre-trained transformer models. This allows for an efficient search mechanism to identify relevant information. The generative AI layer ensures the responses are coherent, contextually aligned, and user-friendly. Designed to cater to diverse user groups, the chatbot is beneficial for researchers, legal professionals, corporate analysts, and healthcare practitioners. It offers robust features such as summarizing lengthy documents into concise overviews, addressing specific user queries with pinpoint accuracy, and ensuring scalability to manage multiple users and large files simultaneously. Additionally, the chatbot’s intuitive interface provides easy accessibility to non-technical users.
Author's Name: R.DEVI, V.RAMYA, T.AKASH, M.SREENIVASAN, B.UBENDRAN
Abstract— Object detection has considerable importance in areas, such as defense and military applications, urban studies, airport surveillance, vessel traffic monitoring, Marketing and transportation infrastructure determination. A lot of attention has been associated with Machine Learning, specifically neural networks such as the Convolutional Neural Network (CNN) winning object Detection. Need to focus on typical generic object detection architectures along with some modifications and useful tricks to improve detection performance further. This paper presents a model for providing the object detection on the given image and video. Object Detection is one of the core problems in Computer Vision field with a large variety of practical applications. The problem precision of object detection is to discover where objects are located in the provided image.
Author's Name: Neeraj Jain, Naman Sharma, Rohit Yadav, Nitin Kumar
Abstract— In today’s digital age, the exponential growth of textual data from diverse sources such as news articles, research papers, social media, and websites has created an urgent need for effective text summarization tools. This paper presents the design, development, and implementation of a Text Summarization Web Application using Streamlit, a lightweight and user-friendly Python framework for rapid web application development. The system leverages advanced transformer-based natural language processing (NLP) models, specifically the bart-large-cnn model, to perform abstractive summarization, generating concise and coherent summaries that capture the essential meaning of the input text. The web application offers three flexible input modes: direct text input, URL-based extraction, and file upload summarization (supporting PDF and TXT formats). To enhance usability, the system allows users to select different summary length levels — Low (200–300 words), Balanced (600–700 words), and High (1000–1200 words) — and incorporates a chunking mechanism that divides long texts into smaller sections for efficient processing within a specified time limit. The app integrates a clean and responsive user interface with dynamic animations, intuitive navigation, and a customizable layout, improving the overall user experience. Experimental results demonstrate the application’s capability to produce high-quality summaries across various content types, maintaining fluency, informativeness, and coherence. The proposed system significantly reduces the time and effort required to comprehend large documents, making it a valuable tool for researchers, students, content creators, journalists, and professionals across domains. This work also highlights the potential of integrating cutting-edge NLP models into accessible web applications to bridge the gap between complex AI technologies and end users. Future work will focus on expanding language support, incorporating extractive summarization options, implementing domain-specific fine-tuning, and enhancing scalability through cloud deployment.
Author's Name: Arvind Sharma, Ansh Tomar, Harit, Hasim
Abstract—With the rapid expansion of online food delivery platforms, accurately predicting customer ordering behavior has become increasingly important for optimizing operations, reducing delivery times, and enhancing user satisfaction. The proposed Online Food Order Prediction System aims to address this challenge by leveraging historical ordering data and user behavioral patterns to forecast future food orders effectively. This system utilizes a combination of machine learning algorithms and key features such as order history, time of day, user preferences, weather conditions, and location to generate real-time predictions. The application processes structured and unstructured data to build predictive models capable of identifying trends and anticipating customer needs. These predictions can be used by food delivery services to optimize resource allocation, improve restaurant inventory management, and personalize marketing efforts. The system is designed to be scalable and adaptable, allowing integration with existing food delivery platforms and point-of-sale systems. To ensure accuracy and minimize overfitting, the model incorporates techniques such as feature engineering, cross-validation, and real-time feedback loops. Additionally, the system supports a user-friendly interface for visualization and interpretation of predictions, enabling actionable insights for businesses and users alike. By harnessing the power of data analytics and AI, the proposed solution demonstrates the potential to transform food delivery logistics and enhance customer experience. Future developments may include deep learning models for personalized recommendations, real-time demand forecasting, and integration with smart kitchen and delivery infrastructure.
Author's Name:S.Pavisya dharshini, M.Revathi , R.Suvetha
Abstract— The population of India recorded in 2014 was 1266.26 million and estimated to record 1443.03 by end of 2024. India follows with one of the largest road network after United States spanning over 5.89 million Km as per Industry report. Such huge population even seek for proper management in case of parking at cumbersome or heavy traffic zones. This project deals with one such case in Bhopal capital of Madhya Pradesh. The identified location is situated in MP nagar zone II surrounded with commercial buildings, coaching institutions, Habibgang Railway Station and Proposed Metro Station which is now under construction. All these reasons lead to heavy traffic entire day. Multi level Smart parking is proposed with a structure G+4 and the designing of the structure is done using computer aided applications namely Autodesk Autocadd and Autodesk 3d Max and analyzed the structure considering seismic zone as per I.S. 1893-I:2016 using Bentley Staad.Pro.
Author's Name: Neeraj Jain, Rahul, Ronak Gurjar, Sagar Yadhuvanshi, Mohit Yadav, Chetan Yadav
Abstract— The rapid advancement of artificial intelligence and smart technologies has opened new possibilities in the field of education. This paper presents the concept of a Smart Teacher system designed to enhance the teaching and learning experience through intelligent automation and data-driven insights. The Smart Teacher integrates technologies such as machine learning, natural language processing, and computer vision to monitor student engagement, personalize content delivery, and provide real-time feedback to both students and educators. By analyzing student behavior, performance, and interaction patterns, the system can adapt teaching strategies to meet individual needs and improve learning outcomes. The Smart Teacher acts as a supportive tool for educators, aiming to create a more interactive, inclusive, and efficient classroom environment.
Author's Name: Koustubh Menon, Minoliya , Shubhi
Abstract— The Primitive usage of URL is to use as a web address. Some URLs can be used to host unsolicited content that can potentially result in cyber attacks. These URLs are called malicious URLs. The inability of the end user system to detect and remove the malicious URLs can put the legitimate user in vulnerable condition. Furthermore, usage of malicious URLs may lead to illegitimate access to the user data by adversary. The main motive for malicious URL detection is that they provide an attack surface to the adversary. Malicious URL detector is a system which is mainly proposed to eliminate unwanted websites which affects online privacy and system health. Nowadays the world is revolving with full of internet activites so the dark web activites also increased which may be more harmful for today’s youngsters and children which may affect their life cycle and privacy so in order to prevent these dark web activities we propose a solution with high integrated classification model by gathering web parameters and analysing it with dataset model. The system takes the web link as a input and scans for any malicious or malware contents inside it and alerts the user.
Author's Name: J. R. Arun Kumar, Samendra Singh Naruka, Gurjot Singh, Mahaveer Singh
Abstract—Image colorization is the process of automatically adding colors to grayscale images to enhance their visual appeal and interpretability. Traditionally, this task required manual input or rule-based algorithms that lacked the ability to generalize well to diverse image features. With the rise of deep learning, particularly Convolutional Neural Networks (CNNs), image colorization has become significantly more accurate and efficient. This project presents a CNN-based approach for converting grayscale images into color by learning spatial and contextual features from large datasets. The model is trained on RGB images converted to grayscale, where it learns to predict the chrominance components (a and b channels) of the LAB color space while using the luminance (L channel) as input. The network architecture is composed of several convolutional layers designed to extract low- to high-level features, enabling the model to produce colorized outputs that are both realistic and visually appealing. The system is implemented using Python and TensorFlow, and it is trained and evaluated on standard datasets like CIFAR-10 and ImageNet subsets. The results show that the model is capable of restoring plausible color information in grayscale images, outperforming basic colorization methods. Potential applications of this project include the restoration of old black-and-white photographs, video frame colorization, enhancement of medical imaging, and creative image editing. Future improvements may involve integrating advanced models such as Generative Adversarial Networks (GANs) or incorporating attention mechanisms to further refine color accuracy and enhance image detail.
Author's Name: G.R.V Lakshmi, B.V.N.S. Sai Saran, A. Narendra, B.S.V.G.V.N. Raju
Abstract— Edge detection is a basic step in image segmentation; it helps to extract the foreground pixels from background especially in medical images.The main purpose of the work is to develop the edge detection on medical images. This method can be used for medical diagnosis in the department of radiology. The existing systems have provided sufficient information to achieve the edge detection, even though the results have to be improved. This paper presents computer vision system with grey scale images for the edge detection. In this work a modified Moore neighbourhood edge detection algorithm is used on medical images. The main goal of this work is to detect the boundaries of medical images with Moore neighbourhood edge detection algorithm to help for medical diagnosis.
Author's Name: Mohit Sharma, Harsh Sharma, Devanshu Saini, Ankit Singh Naruka, Krishan kant. Kamal Kant
Abstract— In the digital age, security and authenticity are critical to maintaining the integrity of voting systems. This paper proposes a Face Recognition Voting System specifically designed for college elections. The system leverages facial biometrics to authenticate voters, preventing fraudulent activities such as duplicate voting and identity theft. Built using machine learning and computer vision algorithms, the system enables real-time face verification with a user-friendly interface for both voters and administrators. The primary aim is to enhance the transparency and reliability of campus elections.
Author's Name: : Dipyan Kumar Ghosh, Parvindar Singh, Tanmay Yadav, Shoheb Akhtar, Puneet Moulakhi, Sandeep
Abstract— IIn the realm of inclusive technology, communication barriers faced by the deaf and hard-of-hearing community continue to present significant challenges. The "Sign Language Detection System" project aims to bridge this gap by providing a real-time, AI-powered web-based tool that detects hand gestures and translates them into corresponding text. Utilizing Python, MediaPipe for hand tracking, and a trained machine learning model, the system accurately identifies static ASL gestures and displays the results on a lightweight interface. Integrated with a webcam-based live feed and designed with a responsive GUI using HTML, CSS, and JavaScript, the system offers a practical and accessible solution for basic sign-to-text translation. This paper explores the design, implementation, and potential improvements of the system as a modern assistive technology for communication enhancement.
Author's Name: : Rashmi, Nalina
Abstract— Internet of Things has become one of the booming technologies in the recent decades and has a major contribution in making our lives more efficient. This paper aims at automating all the electrical appliances in our house and monitoring their status through an Android device using Raspberry Pi. And also security is provided through a GSM Module. This project uses few IoT related algorithms to monitor the status of electrical appliances and a servo motor which acts as door and is unlocked when the authorized user enter appropriate commands in his Android device..
Author's Name: : Mohit Sharma, Keshav Saini , Aryan Kumar , Kartikey Sharma , Dipesh Sirohia
Abstract— The advancement of Artificial Intelligence (AI) and Natural Language Processing (NLP) has significantly transformed the way humans interact with machines. Conversational AI systems, such as chatbots, have become increasingly popular in various domains including customer service, education, and personal assistance. This project, titled NOVA – AI Conversational Chatbot, focuses on the design and development of an intelligent chatbot capable of engaging in meaningful and human-like conversations with users. NOVA is built using a combination of frontend and backend technologies, integrated with an AI-powered language model to interpret and respond to user inputs. The chatbot aims to provide relevant and context-aware answers, making interactions smoother and more intuitive. The system is designed to be scalable, customizable, and adaptable for use in different environments, such as websites, applications, and informational platforms. This journal outlines the project's objectives, architecture, implementation process, and potential applications, highlighting the scope and impact of conversational AI in modern digital systems.
Author's Name: Spoorthi, Sunitha
Abstract— SMART ERP System is a system that deals with colleges` event management. As we can find that nowadays various events are being organized on college-level either technical or non-technical. Various systems are developed for organizing events. Different systems focus on different features like some have a focus on a variety of events to be organized, some focus on personalization of dashboards, and other features like a feedback system for both students and teachers, certification record, student registration record, fee payment system. But still, various organizations are unable to find or develop a system suitable as per their requirements or which contains all the features in one place and this is the reason that various institutions are still doing mostly tasks manually. The main objective is to create an event management site that can organize a variety of events(like workshops, training, competitions, etc), all in one place. The admin or project coordinator can add any event with image and description. The teachers and admin are also allowed to update and delete the events. The student can register in any technical or non-technical event online itself. It also provides personal dashboards for faculties and students, to keep track of their participation in the events. Provide a feedback system for both students and teachers, certification record system, and student registration record system. To automate the fee payment system. To reduce the manual efforts done by the students and teachers. To make the process of organizing and participating in events a much easier task.
Author's Name: Rajesh Kumar, Raju Kumar Pandey, Dhiraj Kumar Rawani, Jitu Pandey
Abstract— In today’s digital world, secure data communication is a critical challenge due to the ever-increasing cyber threats. This paper presents a digital steganography-based approach for secure data hiding within digital media such as images, audio, and text files. The project introduces a system using Java and Spring Boot that implements audio, image, and text steganography through Least Significant Bit (LSB) techniques and text-based methods. This approach ensures that sensitive information remains imperceptible and secure during transmission. The system also integrates email notifications for user alerts and offers a simple interface for encoding and decoding hidden data, highlighting potential real-world applications in secure communication.
Author's Name: Pooja Singh, Pranjal Kushwaha
Abstract— On the surface, it may look like a normal PC controls the machines, but the computer's unique software and control console are what really sets the system apart for use in CNC machining. Under CNC Machining, machine tools function through numerical control. A computer program is customized for an object and the machines are programmed with CNC machining language (called G-code) that essentially controls all features like feed rate, coordination, location and speeds. With CNC machining, the computer can control exact positioning and velocity. CNC machining is used in manufacturing both metal and plastic parts.Inspiring from this CNC technology and revolutionary change in the world of digital electronics &Microcontroller, we are presenting here an idea of “Arduino Based CNC Machine Controller”. The idea behind this concept is to make a small Two Axis CNC router which can engrave 2D & Gray scaled images or pictures with help of high watt burning laser module on surface which can be a paper, wood, leather, plastic, foam etc. It uses two stepper motors as linear actuators on each axis X, Y. While engraving, the proper synchronizationof all this axis i.e. stepper motors, is most challenging task.
Author's Name: Rajesh Kumar, Akash Saini, Parth Jhalani
Abstract— With the advent of social media, people are now more comfortable than ever to express their thoughts, opinions, and emotions online. The proliferation of these comments, whether positive or negative, makes it crucial to analyse them accurately to grasp the true intentions of the writer. To achieve this, sentiment analysis is used to decipher the perspective of the text. In our study, we propose a novel approach that considers the sentimental aspects of the item being reviewed. To validate our approach, we utilized Amazon consumer reviews, specifically the Amazon musical Instruments Reviews dataset collected from the Kaggle repository by Eswar Chand. In this dataset, user ratings were initially detected in each analysis, after which we conducted pre-processing operations, such as creating a sentiment column, tokenization, reviewing textpunctuation cleaning, and eliminating stop-words to extract meaningful information such as the positivity or negativity of the feedback. Our main goal was to analyse this data on an aspect level, which would be highly beneficial to marketers in comprehending consumer preferences and adapting their strategies accordingly. Furthermore, we also provide insights into possible future work for text classification. Ultimately, our study presents a new approach to sentiment analysis that can enhance our understanding of online feedback and facilitate more effective marketing practices.
Author's Name: Dipayan Kumar Ghosh, Prince Gupta, Neeraj Saini, Nitin Saini, Mohit Saini, Yuvraj Sain
Abstract— A School Management System is a comprehensive platform developed to streamline and digitize academic and administrative tasks in educational institutions. With the increasing complexity of managing student data, class schedules, staff records, and academic performance, manual processes often lead to inefficiencies and errors. The COVID-19 pandemic further emphasized the need for digital transformation in schools. This system automates core functions such as attendance tracking, result management, timetable generation, and communication between stakeholders, significantly improving productivity and data accuracy. The proposed School Management System utilizes web technologies including PHP, HTML, CSS, JavaScript, and MySQL to build an interactive and user-friendly platform. By incorporating a modular architecture and relational database management, the system enables real-time access to vital information for administrators, teachers, and students. The automation reduces workload, minimizes human error, and secures sensitive educational data. Evaluation from a real-use case in a secondary school environment indicates notable improvements in efficiency and transparency in school operations.
Author's Name: Kaustubh Shetake, Viresh Gaddi, Vaishnavi Shetake
Abstract— ML-kNN is a well-known algorithm for multi-label classification. Multi-label classification has more frequently used in recent years. Although it is more useful in some cases, ML-kNN has major issues due to the fact that it is a binary relevance classifier which only takes one label every time. In this paper, we proposed a lazy learning approaches to classify an unseen instance on the basis of its k nearest neighbors to solve the multi-label classification problem . We collect different real-word data sets from various domains for the experiment. By introducing the coupled similarity between class labels, the proposed method utilize the correlations between class labels, which overcomes the shortcoming of ML-kNN. Experiments on standard data sets show that our proposed Coupled Multi-Label k Nearest Neighbor algorithm (CML-kNN) reachs heigher performance than some existing multi-label classification algorithms. We believe that nearly utilizing k-nearest neighbors is useful to solve the multi-label problem.
Author's Name: Mohit Sharma, Prince Kumar, Ritik Choudhary, Kuldeep Yadav
Abstract—The rapid advance This paper explores the synergistic integration of Natural Language Processing (NLP) and Automatic Speech Recognition (ASR) within a chatbot framework. It investigates methodologies for enhancing conversational agents by enabling seamless voice interaction and leveraging NLP techniques for improved understanding, intent recognition, and response generation from spoken input. The study details the design and implementation of such a system, evaluating its performance in terms of accuracy, latency, and user experience. The findings highlight the potential of this integrated approach to create more intuitive and accessible human-computer interactions.
Author's Name: Ranvir Mahul Singh
Abstract— Now a day, when people visit an unknown/unfamiliar place, it becomes quite difficult for that person to find a local guide and even if we get, we have to pay for it. Sometimes, we have been charged relatively high if we don’t know how to deal with it. To address the above-mentioned problem, we have come up with an idea to develop an application that simulates a local guide. The application mainly provides a set of probable services that could have been able to provide by a local guide. The application mainly includes features like providing information related to popular places and how to visit this places it also provides information related to hangout places, shopping places, monuments, restaurants and also the facts about the place another feature that is also included is the most popular cities and also the information related to it we also have a feature called attributes and friends and features like weather forecast compass checklist are also included.
Author's Name: Rajesh Kumar, Dev Prakash Mishra, Aanand kumar, Kamal Kumar Sharma, Rehan Khan, Utkarsh Sisodia
Abstract— AI & Ayurveda for Smart Animal Healthcare is an intelligent platform that leverages machine learning and traditional Ayurvedic knowledge to assist pet and livestock owners in identifying and treating animal diseases. The system allows users to upload images of affected areas or select symptoms based on observed animal behavior (for common species like cows, dogs, and cats). A trained ML image classification model analyzes the uploaded image to predict potential health issues, while a symptom-matching engine suggests diseases and corresponding Ayurvedic or home remedies. Built using Django, HTML, CSS, and JavaScript, Ayurvet bridges the gap between modern AI-driven diagnostics and eco-friendly healthcare solutions, particularly in rural or underserved areas. By integrating AI for disease detection, rule-based diagnosis, and natural treatment recommendations, Ayurvet empowers users to act quickly, reducing treatment time and cost. Future plans include expanding the dataset for more animal species, voice input, mobile access, and real-time vet consultations.
Author's Name: Shrivitthal Biradar, Khushal Tiwari, Swapnil Pandit, Priyanka Waghmare
Abstract— Project scheduling is very essential part of the project planning phase. Cost optimization of project can be achieved by completing the project before the completion time of the project, which saves additional overhead cost of the construction project. To complete the project work before the scheduled time activity crashing is the best solution to that. Project having more activities can be difficult and so linear programming technique can be applied to the project crashing work and the optimum value to crash the project duration for the desired time can be calculated using LINDO (linear programming software). Three different options are selected having different construction materials for a defined size of floor work. Crashing of project is carried out for a commercial building 4192 sq ft super built up area.
Author's Name: Mohit Sharma, Siddharth Adhalakha, Ajay Sharma, Dheeraj Baylaan
Abstract— The rapid advance Premvati is an interactive e-commerce platform dedicated to plant enthusiasts and gardening lovers, offering a seamless online experience to explore, purchase, and learn about a wide variety of plants. Addressing the challenges faced by users in accessing quality plants and gardening products locally, the platform integrates a user-friendly interface with a curated catalog to support both novice and experienced gardeners. Built using modern web technologies including React.js for the frontend and Node.js for the backend, Premvati ensures a responsive, scalable, and secure environment for online plant shopping. Key features include personalized plant recommendations, care tips, and an intuitive shopping experience tailored to users' preferences and environmental conditions. By combining technology with nature, Premvati aims to foster a greener lifestyle and build a vibrant community of plant lovers. Future developments include integration of plant care reminders, expert consultations, and mobile app support to further enrich the user experience.
