Current Issue

Area of research: Electronics And Communication Engineering

Brain Tumors Result From The Abnormal And Uncontrolled Growth Of Cells. If Left Untreated During The Early Stages, They Can Become Life-threatening. Although Numerous Significant Advancements Have Been Achieved In This Field, Ensuring Accurate Segmentation And Classification Remains A Complex Challenge. The Primary Difficulty In Detecting Brain Tumors Lies In The Variations In Their Location, Size, And Shape. This Paper Aims To Provide An Extensive Review Of Brain Tumor Detection Methods Using Magnetic Resonance Imaging (MRI) To Support Researchers In Their Work. It Encompasses Discussions On The Structure Of Brain Tumors, Publicly Accessible Datasets, Image Enhancement Techniques, Segmentation Methods, Feature Extraction, Classification Approaches, And The Role Of Advanced Technologies Such As Deep Learning, Transfer Learning, And Quantum Machine Learning In Analyzing Brain Tumors. Lastly, This Survey Summarizes Key Findings, Highlighting The Advantages, Limitations, Advancements, And Potential Future Directions In Brain Tumor Detection Research.

Author: Anisha Banu A | Deepa Mathi R | Nutheti Likhitha Chowdary

Area of research: Civil Engineering

Organic Waste Management Through Composting Presents A Sustainable Method To Reduce Landfill Dependency, Mitigate Greenhouse Gas Emissions, And Produce Nutrient-rich Soil Amendments. This Study Investigates The Efficiency Of Composting Processes With And Without The Use Of Accelerators By Analysing Critical Parameters Such As Temperature, Moisture Content, PH, And Carbon-to-nitrogen (C:N) Ratio. Over A 55-day Period, Two Compost Piles—one With An Accelerator And One Without—were Monitored To Evaluate Decomposition Dynamics. Results Demonstrate That The Addition Of Accelerators Enhances Microbial Activity, Accelerates Decomposition, And Produces Mature Compost More Rapidly. This Research Provides Actionable Insights Into Optimizing Composting Processes In Both Small-scale And Large-scale Applications While Emphasizing Best Practices For Achieving High-quality Compost.

Author: Susindiran S | Roopa D

Area of research: CSE

The Agriculture Management System (AMS) Is A Smart, Machine Learning-based Platform That Assists Farmers In Making Data-driven Decisions For Crop Selection, Fertilizer Use, Irrigation Planning, And Yield Estimation. It Analyzes Key Factors Such As Soil Nutrients (N, P, K), Temperature, Humidity, PH, Rainfall, And Historical Yield Data To Deliver Personalized Recommendations. Featuring A Responsive Bootstrap 4 Interface, AMS Ensures Smooth Access Across Devices, Allowing Users To Input Real-time, Location-specific Data For Tailored Insights That Enhance Productivity And Resource Efficiency. The System Integrates Weather APIs For Dynamic, Context-aware Guidance, Helping Farmers Adapt Practices To Current And Forecasted Conditions. A Built-in Agriculture Chatbot Provides 24/7 Support On Topics Like Pest Control, Organic Farming, And Crop Health. An Intelligent Irrigation Calendarfurther Optimizes Water Use By Generating Schedules Based On Crop Type, Soil, And Local Weather. Additionally, The System Stores User Data Securely, Enabling Farmers To Track Their Seasonal Progress And Refine Strategies Over Time. It Supports Multilingual Interfaces To Reach Farmers Across Diverse Regions. The Modular Design Also Allows For Future Integration With Government Schemes And Market Price Updates. In Essence, AMS Empowers Modern Agriculture By Combining AI, Real-time Data, And Intuitive Design To Boost Efficiency And Support Informed Farming Decisions.

Author: Mrs.V.Hemalatha | N.Rejiya Sulthana | S.Priya | A.Pavithra

Area of research: Electronics And Communication Engineering

Congenital Heart Diseases (CHDs) Are Among The Leading Causes Of Mortality Worldwide, Necessitating Early And Accurate Detection Methods.Improving Patient Outcomes And Facilitating Prompt Medical Intervention Depend Heavily On Early Diagnosis. Conventional Heart Sound Analysis Depends On Skilled Medical Professionals Performing Manual Auscultation, Which Is Subject To Human Error And Subject To Subjectivity. Automated Heart Sound Classification Has Been Made Possible By Recent Developments In Deep Learning And Artificial Intelligence (AI), Which Improve Accuracy And Lessen Reliance On Manual Diagnostics. Recurrent Neural Networks (RNN) And Long Short-term Memory (LSTM) Networks Are Used In This Study's AI-powered Heartbeat Sound Analysis System To Accurately Classify Heart Sounds. In Order To Differentiate Between Normal And Abnormal Patterns, The System Is Trained On Phonocardiogram (PCG) Recordings, Which Capture Temporal Dependencies In Heartbeats. The System Is Integrated With Django, A Web-based Framework That Makes It Easier To Process, Store, And Visualize Heart Sound Recordings In Real Time, In Order To Improve Accessibility And Usability. Patients And Medical Professionals Can Effectively Monitor Heart Health Thanks To This Smooth Integration. In Addition To Increasing Diagnostic Precision, The Suggested System Complies With Legal Requirements Like HIPAA And GDPR, Guaranteeing The Security And Privacy Of Patient Data. Even In Places With Limited Resources, Early Detection Of Congenital Heart Diseases Is Now Easier Thanks To The Model's Support For Remote Monitoring Through Cloud-based Deployment.

Author: Poojashree S | Shalini S | Monisha R | Dr.D. Arul Kumar

Area of research: Biomedical Engineering

Diabetic Retinopathy (DR) Is A Leading Cause Of Vision Loss Globally, Particularly Among Individuals With Long-standing Diabetes. Early Detection And Grading Of DR Are Vital To Prevent Irreversible Blindness. This Project Presents An End-to-end, AI-powered Web Application For The Automatic Classification Of Diabetic Retinopathy From Fundus Images Using A Convolutional Neural Network (CNN) Deployed Via A FastAPI Framework. The Trained Model, Based On TensorFlow, Classifies Input Retinal Images Into Five Categories: No DR, Mild, Moderate, Severe, And Proliferative DR. The Dataset Used To Train The Model Was Sourced From Kaggle, Consisting Of High-resolution Retina Images Labeled By Clinical Experts. The Application Provides A Modern, Responsive Frontend Using HTML, CSS, And JavaScript, Allowing Users To Upload Retinal Images And Receive Real-time Diagnostic Predictions. The Backend Model Preprocesses Uploaded Images Using OpenCV And NumPy, Resizing Them To 224x224 Pixels, Normalizing Them, And Feeding Them Into The Trained CNN Model. The System Aims To Serve As A Fast, Reliable, And Accessible Tool To Assist Ophthalmologists And Healthcare Professionals In Screening For DR. It Can Also Act As A Valuable Aid In Regions With Limited Access To Medical Infrastructure, Where Regular Eye Checkups Are Not Always Feasible. Through This Project, We Demonstrate The Real-world Application Of AI And Web Development For Medical Diagnostics, Bridging The Gap Between Complex Deep Learning Models And User-friendly Interfaces.

Author: Dhivan T | Illayaraja R | Suresh Gopi B | Dr Mythili S

Area of research: ISE

This Project Introduces A Real-time Fire Detection And Alert System Based On The YOLO Algorithm, Aiming To Enhance Safety In Various Settings. Designed To Mitigate Fire Hazards Across Diverse Environments, Including Industrial, Residential, And Public Spaces. Leveraging The YOLO Deep Learning Algorithm, The System Accurately Identifies Fire From Live Camera Feeds And Triggers Immediate Responses Through Integrated Hardware. On Detecting Fire, It Activates An Alarm, Captures Evidence, Emails Alerts, And Initiates A Water Pump Via An ESP32 Microcontroller. This AI-powered System Eliminates The Need For Traditional, Maintenance-heavy Sensors And Instead Offers A Scalable, Cost-effective Alternative That Integrates Seamlessly With Existing Surveillance Setups. By Ensuring Rapid Detection And Response, It Enhances Environmental Safety And Operational Efficiency.

Author: Chaya P | Anu Priya | Chinmayi D | Lavanya S | Waiza Fathima

Area of research: Mechanical Engineering

This Project Is Related To Transferring Goods From A First Horizontal To Second Horizontal Conveyor Comprising A Substantially Upright Extending Frame- Endless Drive Arranged On The Frame And Drivable By A Motor; At Least One Support Member Which Is Connected To Endless Drive And Which Is Drivable In A Circuit By Means Of The Endless Drive. At Least One Product Carrier Connected To The Support Member, Wherein The Product Carrier Is Connected To The Support Member For Rotation About Lying Shaft Extending Transversely Of The Frame, Wherein The Product Carrier Is Connected Drivably To No More Than Only One Trolley. So, It Is Basically A Vertical Conveyor With A Carriage Which Is Mounted On Endless Chain And It Lift Boxes In Vertical Direction And Dispatch Them On Another Horizontal Conveyor Synchronized With It.

Author: Mrs. J. S. Tilekar | Akshay Dange | Manohar Pawar | Shivraj Rite | Vishwajeet Nimbalkar

Area of research: Computer Science And Engineering

TalkyCommunity.com Is An All-in-one Learning Solution, Providing Holistic Education That Is Interactive. The Program’s Integration Of Social Interaction With Learning Activities Gives Individuals The Chance To Discover New Concepts, Sharpen Existing Skills, And Engage In Educational Discourse. Talky Community Enables Formal And Informal Learning Through Its Multifunctional Interface, Including Discussion Boards, Resource Repositories, Event Calendars, And Peer Networking. This Journal Investigates The Platform’s Impact On Enhancing Equitable Education, Supporting Self-paced Development, And Cultivating A Welcoming Online Environment That Encourages Exploration, Teamwork, And Development. The Platform’s Design Emphasizes User Engagement, Making Learning More Collaborative And Enjoyable. Its Flexible Structure Caters To A Wide Range Of Learning Styles And Educational Backgrounds. By Encouraging Learners To Share Knowledge And Experiences, Talky Community Creates A Vibrant Ecosystem Of Peer Support. It Also Plays A Vital Role In Bridging Educational Gaps By Making Resources Accessible To Diverse Communities. As A Result, Learners Are Empowered Not Only To Absorb Information But Also To Actively Contribute To The Learning Journey Of Others.

Author: E. Sinega | J. Sowmiya | S. Dhulasilinkam | K. M. Anbu | K. Kavipriya

Area of research: Computer Science And Engineering

This Paper Presents A Digital Platform That Serves As A Comprehensive Resource For The Study Of Medicinal Plants Used In Ayurveda (Ayurveda, Yoga And Naturopathy, Unani, Siddha, And Homeopathy). It Provides Detailed Information On The Characteristics And Uses Of These Plants, Aimed At Students, Physicians, And Those Interested In Natural Medicine. The Plat- Form Combines Traditional Knowledge With Modern Technology, Fosters A Deeper Understanding And Appreciation Of Medicinal Plants, And Preserves This Valuable Wisdom For Future Generations

Author: Sanika Algude | Pratik Bevnale | Jatin Bhujbal | Roopam Wadkar | Mansi Bhonsle

Area of research: Mechanical Engineering

This Paper Presents A Railway Crack Detection Robot Using Arduino, Designed To Detect Cracks And Defects In Railway Tracks. The Robot Is Equipped With Sensors That Detect Cracks And Transmit Data To A Central Station. The System Aims To Improve Railway Safety By Identifying Potential Defects Before They Cause Accidents. The Robot's Design And Functionality Are Tailored To Navigate Railway Tracks, Detect Cracks, And Provide Real-time Data To Maintenance Teams. This Project Demonstrates The Potential Of Automation And IoT In Enhancing Railway Safety And Maintenance. In Existing Method Of Crack Detection, We Not Get Closer Location Of Crack We Only Get Location Of That Crack In Format Of Latitude And Longitude. So, This System Is Modified, By Adding Paint Drop System Which Gives Exact Location Of Crack In Format Of Longitude And Latitude And Also Drop Paint Near Crack, So We Get Closer Location Of Crack & For Battery Charge We Add Solar Panel.

Author: Mrs. Jyoti S.Tilekar | Mr.Shreyah Borate | Mr. Karan Jadhav | Mr.Amit Jagtap | Mr. Mohammadrehan Attar

Area of research: Artificial Intelligence / Machine Learning

Navigating Daily Life Can Be Especially Challenging For Blind And Visually Impaired Individuals, Particularly When It Comes To Identifying Obstacles, Recognizing Familiar Faces, And Handling Currency Transactions. Traditional Aids Such As White Canes And Guide Dogs, Though Helpful, Provide Limited Functionalities And Are Not Equipped To Handle Dynamic Environments Or Complex Tasks In Real Time. This Paper Introduces An Innovative Solution That Integrates Face Detection, Obstacle Detection, And Currency Recognition Into A Single, Wearable Device. By Utilizing Cutting-edge Artificial Intelligence, Including The Grassmann Model For Face Recognition, YOLO For Object Detection, And Convolutional Neural Networks (CNNs) For Currency Identification, The Proposed System Empowers Visually Impaired Individuals To Navigate Their Surroundings, Recognize People, And Manage Financial Transactions Independently. Real-time Image Capturing And Processing Allow For Immediate Audio Feedback, Ensuring Users Receive Context-sensitive Assistance As They Encounter Various Challenges In Everyday Settings. This System Not Only Enhances User Autonomy But Also Reduces Dependence On External Assistance, Fostering Greater Confidence And Independence. By Combining Multiple Functionalities In A Single, User-friendly Device, The Proposed Solution Addresses Gaps In Existing Technologies, Offering A Practical And Affordable Alternative For Enhancing The Quality Of Life Of Visually Impaired Individuals.

Author: Hameed Asik K | Naveen Kumar R | Kartheeswaran V | Vimala D

Area of research: Compute Science

An Extensive Web-based Program Called The Employee Management System (EMS) Was Created To Automate And Simplify The Fundamental Tasks Of Human Resource Management In A Company.Effective Management Of Employee-related Tasks, Including Department Allocation, Payroll Processing, Attendance Monitoring, Performance Reviews, And Report Creation, Is Made Possible By This System.Through The Reduction Of Manual Labor, The Mitigation Of Human Error, And The Provision Of Prompt Access To Employee Data, The EMS Increases Productivity.Constructed Utilizing Contemporary Web Technologies Such As HTML, CSS, And JavaScript, The System Offers A Comprehensive Capability And An Intuitive User Experience Specifically Designed For HR Administrators And Management Teams.All Things Considered, The EMS Is Essential To Preserving Correct Personnel Data And Enhancing Organizational Effectiveness.

Author: S.Manishankar | K.Mathesh | P.Ramanathan | S.Renganathan

Area of research: NURSING

Author: Neha Patyal | Anamika Saini | Kajal Banyal | Jasbir Kaur | Kalpna Chauhan

Area of research: Computer Science And Technology

Traditional CAPTCHA Systems, While Effective In Deterring Basic Automated Threats, Often Create A Cumbersome User Experience And Are Increasingly Susceptible To Modern AI- Driven Attacks. This Paper Presents A Machine Learning (ML)- Driven Passive CAPTCHA Alternative Specifically Designed For The Unique Identification Authority Of India (UIDAI) Portals. By Passively Collecting Environmental And Behavioral Data—such As Mouse Dynamics, Keystroke Patterns, Device Fingerprints, And Net- Work Indicators—our Proposed Solution Leverages Backend ML Models To Assess User Authenticity In Real-time. The Architecture Promotes Minimal User Interaction, Seamless Integration With UIDAI Infrastructure, And Robust Security Against DoS/DDoS Threats, All While Upholding Strict Privacy Guidelines.

Author: R Kamal Raj | Gnanavika M | Shreyas D M | Yamanappa

Area of research: Electronics And Communication Engineering

Fault Detection In Industrial Systems Is Crucial For Ensuring Operational Safety, Minimizing Downtime, And Reducing Maintenance Costs. This Work Proposes A Hybrid Deep Learning Model Combining Convolutional Neural Networks (CNN) And Gated Recurrent Units (GRU) To Detect And Classify Machine Faults From Time-series Data. The CNN Layers Extract Spatial Features, While GRU Layers Model Temporal Dependencies In The Data.The Architecture Incorporates Residual Connections To Enhance Gradient Flow And Improve Learning Efficiency. The Model Is Evaluated On Multi-class Fault Detection Datasets, Achieving Robust Performance With High Accuracy, Precision, Recall, And F1-score. Advanced Metrics, Including ROC- AUC, Logarithmic Loss, Cohen's Kappa, And Matthews Correlation Coefficient, Demonstrate The Model's Reliability. Visualization Of Confusion Matrices And Detailed Performance Metrics Validates Its Effectiveness In Detecting Anomalies And Classifying Fault Types. This Approach Can Be Generalized For Real-time Monitoring Systems In Various Industrial Applications, Ensuring Predictive Maintenance And Operational Excellence.

Author: Gnanaprakash J | Gobichandar M | Meenatchi K | Praveena G | JOSEPH S

Area of research: Computer Science

Off Late Conventional Image Data Hiding And Encryption Mechanisms Have Seen A Shift Towards Homomorphic Images Which Can Be Thought Of Being Created From A Constant Illumination And A Varying Reflectance. In This Proposed Work, The Fresnel Transform Is Employed To Convert Normal Images Into Homomorphic Images To Reduce The Redundancy Of Images. Subsequently, The Image Is Converted To The Transform Domain Using The 4th Level Discrete Wavelet Transform. The Truncation Of The DWT Is Done At The 4th Level So As To Limit The Complexity Of The System. Once The Image Is Converted To The Transform Domain, It Is Encrypted Using The Chaotic Baker Map.The Embedded Data Can Be Extracted From The Encrypted Domain Itself Without The Mandatory Necessity Of First Decrypting The Image Thereby Making The Secret Image Extraction Faster And Less Perceptible. The Evaluation Of The Proposed Technique Is Done Based On The Histogram Analysis, The MSE, PSNR, Correlation And Entropy. It Has Been Shown That The Proposed System Performs Better Compared To The Previously Existing Technique In Terms Of The PSNR For The Same Image From The Benchmark USC-SIPI Image Dataset.

Author: Ajay Singh Patel | Prof. Sunil Parihar

Area of research: Civil Engineering

Author: C Sudharsan | V.Indhumathi | A.Abitha

Area of research: Artificial Intelligence And Data Science

This Project Presents A Web-based "Smart Parking" System Designed To Enhance Urban Parking Efficiency Through Intelligent Slot Booking And User-friendly Interaction. The System Features A Responsive Landing Page Highlighting Key Capabilities, Including AI-powered Vacancy Detection, Real-time Space Tracking, License Plate Recognition, And Mobile Accessibility. Users Can Authenticate Via A Secure Login Form And Proceed To An Interactive Slot Booking Interface, Where Available And Booked Slots Are Visually Distinguished. The Booking Page Allows Users To Select A Date, Time, And Parking Slot With Immediate Visual Feedback And Confirmation. The Interface Is Built Using HTML, CSS, And JavaScript, With A Focus On Usability And Clarity, Ensuring Accessibility Across Devices. This Solution Demonstrates A Scalable Foundation For Future Integration With IoT And Backend Technologies For Real-time Parking Management.

Author: Sasidaran.D | Vinithkumar.S | Siva.S | Seenivasn.P

Area of research: Computer Science

Agriculture Is The Backbone Of The Indian Economy. Farmers Often Struggle To Achieve Expected Crop Yields Due To Factors Like Unpredictable Weather, Soil Variability, And Lack Of Advanced Forecasting Systems. This Paper Presents A Machine Learning-based Approach For Recommending Crops Based On Historical Data, Including Soil Type, Rainfall, And Crop Productivity, Using A Decision Tree Classifier. The System Provides Valuable Insights To Farmers By Analyzing Region-specific Datasets And Enhances Their Decision-making Capabilities

Author: Indhreesh.R | Sekar M | Lokesh V | Vignesh Kumar.M

Area of research: CSE

Acute Myocardial Infarction (AMI), Or Heart Attack Is A Severe Condition Caused By Reduced Blood Flow To The Heart. Early Detection Is Crucial To Lower Its Global Impact On Health. This Project Presents CardioCare AI, A Machine Learning Based Model For Predicting And Assessing AMI Risk. By Analyzing Data Like Cholesterol Blood Pressure, Blood Sugar, Smoking Habits, And Family History, It Identifies High Risk Individuals With Great Accuracy. Using Advanced Algorithms Like XGBoost Known For Handling Medical Data Effectively, The System Detects Patterns And Relationships Among Clinical Features. CardioCare AI Focuses On Non- Invasive Data To Ensure Its Accessibility For Widespread Use. It Provides Healthcare Professionals With Actionable Insights For Early Intervention, Enabling Preventive Care And Personalized Treatments. The Model Integrates Predictive Analytics Into Daily Medical Practices To Enhance Diagnostic Speed And Reliability, Addressing Limitations Of Traditional Methods. This Innovative Approach Aims To Improve Patient Outcomes Reduce Healthcare Challenges, And Support Global Cardiovascular Disease Prevention Efforts. By Offering A Scalable Solution, CardioCare AI Represented.

Author: YV Akash | V Deepak | S Gurumoorthy | S Jeeva | R Vijay

Area of research: CSE

An Innovative Indoor Automation System Enhances Energy Efficiency By Combining Real-time Object Detection With Environmental Sensing. Using The YOLO Algorithm In Python And Light/temperature Sensors, It Monitors Human Presence And Room Conditions To Control Lights And Fans. The System Includes Three Modules: Computer Vision For Occupant Detection, Sensors For Environmental Monitoring, And An Automated Control Unit For Appliances. It Ensures Devices Operate Only When Needed, Promoting Energy Efficiency And Supporting Sustainable Living.

Author: Chandana K C | Chandana N S | Hemashri H M | Kumuda L | Lavanya S

Area of research: Information Technology

Dental Diseases Such As Dental Caries, Periodontal Disease, Periapical Lesions, And Bone Loss Are Among The Most Prevalent Oral Health Issues Globally. Early And Accurate Detection Is Critical To Preventing Progression And Ensuring Effective Treatment. Traditional Diagnostic Methods Relying On Manual Inspection Of Radiographs Are Often Time-consuming And Subject To Variability In Interpretation. This Study Presents A Deep Learning-based Approach For The Automated Detection And Classification Of Dental Diseases Using Various Types Of Radiographic Images, Including Panoramic, Periapical, And Bitewing Radiographs. A Custom Convolutional Neural Network (CNN) Architecture, As Well As Pre-trained Models Such As ResNet50 And EfficientNet, Were Trained And Evaluated On A Curated Dataset Comprising Over 5,000 Labeled Radiographs Annotated By Experienced Dental Professionals. Image Preprocessing Techniques, Such As Contrast Enhancement And Noise Reduction, Were Applied To Improve Image Quality And Model Performance. The Models Were Trained To Classify Multiple Disease Conditions, Including Caries, Periodontal Bone Loss, Impacted Teeth, Cysts, And Periapical Abscesses. The Proposed Models Demonstrated High Classification Accuracy, With The Best-performing Model Achieving An Overall Accuracy Of 93.2%, Sensitivity Of 91.5%, And Specificity Of 94.8% Across All Radiograph Types. Cross-validation Confirmed The Model’s Robustness And Generalization Across Different Imaging Conditions And Patient Demographics. Furthermore, Class Activation Mapping (CAM) Was Used To Provide Visual Explanations, Increasing The Interpretability Of The Results And Enhancing Clinical Trust. This Study Confirms The Viability Of Deep Learning Systems As A Reliable Tool For Dental Disease Diagnosis. By Leveraging Multiple Radiographic Modalities, The System Enhances Diagnostic Accuracy And Can Serve As A Valuable Adjunct In Clinical Workflows, Reducing Diagnostic Delays And Improving Patient Outcomes.

Author: Darshini N | Dhanusree K | SriVarthini K | Swetha E

Area of research: Mechanical Engineering

The Term “minor Losses”, Used In Many Textbooks For Head Loss Across Fittings, Can Be Misleading Since These Losses Can Be A Large Fraction Of The Total Loss In A Pipe System. In Fact, In A Pipe System With Many Fittings And Valves, The Minor Losses Can Be Greater Than The Major (friction) Losses. Thus, An Accurate K Value For All Fittings And Valves In A Pipe System Is Necessary To Predict The Actual Head Loss Across The Pipe System. K Values Assist Engineers In Totaling All Of The Minor Losses By Multiplying The Sum Of The K Values By The Velocity Head To Quickly Determine The Total Head Loss Due To All Fittings. Knowing The K Value For Each Fitting Enables Engineers To Use The Proper Fitting When Designing An Efficient Piping System That Can Minimize The Head Loss And Maximize The Flow Rate. The Objective Of This Experiment Is To Determine The Loss Coefficient (K) For A Range Of Pipe Fittings, Including Several Bends, A Contraction, An Enlargement, And Agate Valve. In Our Capstone Project Work Our Main Objective Is To Replace The Manometer Which Is Available In The Experimental Set Up By Pressure Gauges For The Accurate Reading And Further Calculations. So That The Effectiveness And Reliability In The Performance Of The Test Ring Must Be Improved In Comparison With The Manometer.

Author: Mr. J. P. Pinjar | Mr .P. P. Mulade | Mr .N. V. Bahirgonde | Mr.P. S. Talbhandare | Mr.M. A. Puppal

Area of research: Artificial Intelligence And Data Science

This Project Looks At How To Create An Engaging Single-player Game Stressing Smooth Controls And Real-time Interaction Developed In Unity Game Engine. A Responsive Gaming Experience Is Possible For Players Who Can Move Smoothly Between Various States Including Walking, Running, Crouching, And Standing Still. The Game Features A Countdown Timer Indicating The Conclusion Of A Match And A Restart Choice, As Well As Systems For Managing Enemy Spawning And Scoring. Weapons Make Fight To Seem Dynamic And Varied By Allowing Both Semi-automatic And Full-automatic Shooting Modes. This Research Proposes A Real-time Speech Emotion Recognition (SER) System That Classifies Human Emotions From Audio Input Using A Machine Learning Pipeline. The System Utilizes The RAVDESS Dataset For Training And Extracts Acoustic Features Such As Mel Frequency Cepstral Coefficients (MFCC), Chroma And Spectral Contrast. A Multilayer Perceptron (MLP) Classifier Is Used For Emotion Prediction, Recognizing Eight Distinct Emotional States. Real-time Audio Can Be Recorded And Analyzed By The Model, Which Adaptively Improves Itself Using High Confidence Predictions. The Proposed System Is Capable Of Dynamic Learning, Thus Continuously Enhancing Performance Over Time. This Approach Facilitates The Integration Of Emotional Intelligence In Applications Such As Virtual Assistants ,mental Health Monitoring, And Interactive Voice-based Systems.

Author: Vasudevan S | Sathishkumar K | Sampathkumar K | Sachin S

Area of research: Information Technology

This Project Builds Upon Existing Research In Plant Disease Classification By Introducing Key Innovations That Enhance Model Efficiency, Accuracy, And Interpretability. While The Document Outlines A Model Based On EfficientNetB3, This Implementation Leverages DenseNet121, Which Improves Feature Reuse, Reduces Overfitting, And Requires Fewer Parameters. Additionally, Extensive Data Augmentation Techniques Such As Random Flipping, Edge Detection, Convolutional Filtering, And Blurring Enhance The Model’s Ability To Generalize Across Diverse Real-world Conditions.

Author: Dr.K.Geetha | Bhuvaneshwaran G | Aravindhan K | Syed AhamedB.A

Area of research: Electrical Engineering

This Paper Explores The Design And Implementation Of An Energy Management System (EMS) For Electric Vehicle (EV) Charging Stations, Utilizing Microcontroller Technology To Optimize Energy Usage. The Proposed System Integrates Renewable Energy Sources, Dynamically Adjusts Charging Schedules Based On Real-time Energy Availability And Demand, And Ensures Efficient Energy Utilization. By Addressing Challenges Like Peak Energy Demand And Renewable Energy Variability, The EMS Enhances The Sustainability And Reliability Of EV Charging Infrastructure. Simulation And Experimental Results Demonstrate The Effectiveness Of This System In Achieving Energy Efficiency And Reducing Environmental Impact.

Author: Ajay Saanjay Pawar | Ashwin Santosh Sonawane | Bhushan Hari Jadhav | Prof. Y. R. Patni

Area of research: Game Development

This Project Looks At How To Create An Engaging Single-player Game Stressing Smooth Controls And Real-time Interaction Developed In Unity Game Engine. A Responsive Gaming Experience Is Possible For Players Who Can Move Smoothly Between Various States Including Walking, Running, Crouching, And Standing Still. The Game Features A Countdown Timer Indicating The Conclusion Of A Match And A Restart Choice, As Well As Systems For Managing Enemy Spawning And Scoring. Weapons Make Fight To Seem Dynamic And Varied By Allowing Both Semi-automatic And Full-automatic Shooting Modes. An Easy To Understand And Friendly User Interfae Lets Players Know With Live Updates On Their Score And Left Time. All Things Considered, The Game Offers An Engaging And Immersive Experience Totally Created In Unity By Combining Good Mechanics With A Simple UI.

Author: Prof. (Ms). Arti Sondawale | Mrunal Wankhede

Area of research: Civil Engineering

Clarifiers Play A Crucial Role In Water And Wastewater Treatment Plants, And Their Failure During Earthquakes Can Pose Significant Threats To Public Health And The Environment. Therefore, Ensuring Their Seismic Resilience Is Essential For Maintaining Operational Functionality During Such Events. This Review Analyzes How Seismic Activity Affects Clarifier Structures, With A Focus On Structural Design Across Different Seismic Risk Zones. It Investigates The Dynamic Forces—including Both Lateral And Vertical Loads—that Act On These Structures During Earthquakes And The Resulting Engineering Challenges. The Review Outlines Effective Design Strategies Aimed At Improving Performance Under Seismic Stress. Clarifiers Typically Consist Of Two Main Functional Zones: The Clarification Zone, Where Sedimentation By Gravity Occurs, And The Thickening Zone, Where Solids Settle And Form A Concentrated Sludge Blanket. The Review Concludes By Underscoring The Importance Of Ongoing Research To Improve The Seismic Durability Of Clarifiers And Other Essential Components Of Water Infrastructure In Earthquake-prone Areas

Author: Mr. Shubham Babasaheb Hatte | Dr. S. A. Bhalchandra

Area of research: Health Science

Sign Language Is A Crucial Communication Medium For Individuals With Hearing And Speech Impairments. However, The Lack Of Widespread Accessibility To Sign Language Interpreter’s Limits Communication Opportunities For The Deaf And Mute Community. This Project Presents A Sign Language Classification And Voice Output System Using ResNet, A Deep Learning-based Model Designed For Accurate Sign Language Recognition. The System Processes Images And Video Frames Of Hand Gestures, Classifies Them Into Meaningful Words Or Letters, And Converts Them Into Speech Output. By Leveraging Convolutional Neural Networks (CNNs) With ResNet Architecture, This System Improves Recognition Accuracy And Real-time Responsiveness. The Model Is Trained Using Benchmark Sign Language Datasets And Optimized With Image Pre-processing Techniques. Performance Evaluation Is Carried Out Using Standard Metrics Such As Accuracy, Precision, Recall, And F1-score. This Study Demonstrates How Deep Learning Can Bridge The Communication Gap For Hearing-impaired Individuals, Providing An Effective Real-time Sign Language Recognition System.

Author: Miss. K.Lalithavani | Deepa.T | Dhivyalakshmi.J | Sahana.R | Sindhu.K

Area of research: Artificial Intelligence And Data Science

We Propose An Intelligent, Automated Seminar Hall Booking System Integrated With A Chatbot Assistant To Streamline Scheduling On Campus. The System’s Objective Is To Replace Manual Reservation Processes—spreadsheets, Emails, And Walk-ins—with A Unified Platform That Handles Booking Requests, Conflict Detection, And User Queries. Technically, We Employ A Three-tier Stack: A Python-based Backend For Booking Logic And NLP Processing, A Java Component For System Integration And Performance-critical Tasks, And HTML/CSS (with JavaScript) For The Responsive Web Interface. The Chatbot Uses NLP Techniques To Understand Natural-language Requests (e.g. “Book A Hall For 50 People Next Tuesday”) And Guides Users Through Available Slots. Automatic Conflict Checking Against The Database Ensures No Double-bookings, And The Calendar Interface Provides Real-time Availability. By Leveraging Automation And AI, The System Is Expected To Significantly Reduce Scheduling Conflicts And Administrative Workload, While Improving User Satisfaction Through 24/7 Conversational Support. In A “smart Campus” Context, This Integrated Approach Enhances Resource Utilization And Transparency: Notifications Keep Students And Staff Informed, And Logged Data Can Inform Future Campus Planning. Early Tests And User Feedback Suggest The Solution Makes Hall Reservation Faster, More Accurate, And More User-friendly.

Author: Nithish Kumar S | Pravin V | Killivalavan T

Area of research: Electrical And Electronics Engineering

The Detection Of Faults In Underground Cables Is Crucial For Maintaining The Reliability And Efficiency Of Electrical Power Distribution Systems. Traditional Methods Of Detecting Faults In Underground Cables Often Involve Manual Inspection, Which Is Time-consuming, Costly, And Inefficient. To Address These Challenges, The Integration Of Internet Of Things (IoT) Technologies Offers A Promising Solution For Real-time Monitoring And Fault Detection. Unist Embedded Systems Have Taken The Initiative To Design And Develop A Comprehensive IoT-based Underground Cable Fault Detection System, Which Allows For Continuous Monitoring And Immediate Identification Of Cable Issues. This IoTbased System Involves The Installation Of Sensors At Strategic Locations Along The Underground Cable Network. The System Uses Advanced Diagnostic Techniques Such As Impedance Measurement And Temperature Analysis To Detect Faults Like Cable Insulation Failure, Short Circuits, And Overheating, Which Can Lead To Failures If Left Unaddressed. The Real-time Monitoring Capability Of The IoT System Enhances The Reliability Of The Cable Network By Providing Timely Alerts Whenever A Fault Occurs. The System Is Capable Of Pinpointing The Exact Location Of The Fault, Significantly Reducing The Time And Cost Involved In Locating And Repairing The Problem. Additionally, The System's Remote Monitoring Feature Allows Maintenance Personnel To Respond Quickly To Issues,thus Preventing Power Outages Or Further Damage To The Network.

Author: Mr.P.Gopinathan | Anupriya.S | Madhumitha.S | Sandhiya.A | Selvalakshmi.G

Area of research: CSE

There Is No Doubt About The Revolutionary Effects Of Artificial Intelligence (AI) On A Variety Of Sectors, Most Notably Research And Education. The Incorporation Of AI Into Library Procedures Has Become An Unavoidable Step, Given Its Significance And The Requirement Of Maintaining Global Competitiveness. In Order To Give A Thorough Overview Of This Dynamic Field, The Paper Primarily Focuses On Explaining How AI-driven Tools And Techniques Are Used In Various Aspects Of Library Operations. Creating Machines That Can Perform Cognitive Tasks Similar To Those Of The Human Brain Is The Main Goal Of Artificial Intelligence. Libraries Can Overcome Physical Limitations And Become More Intelligent And Accessible By Integrating Artificial Intelligence In A Seamless Manner. This Article Explores How The Various Concepts Like Natural Language Processing (NLP), Large Language Model (LLM), Expert System (ES), AI-Powered Indexing Tools, Chatbots And Other AI Tools And Techniques May Change Library Infrastructure And Services In The Future, With The Potential To Improve Outcomes For Students, Teachers, Researchers, And Readers Alike. An In-depth Analysis Of The Benefits, Drawbacks, And Creative Uses Of AI Tools And Technology In Libraries Advances A Comprehensive Knowledge Of The Field And Opens The Door To Wise Decision-making In The Dynamic Field Of Library Sciences.

Author: Ms.K.Abinaya | Ms.k.jayasri | Ms.R.Akalya | Ms.D.kiruthika

Area of research: Information Technology

Hospital Readmissions Are A Major Concern In Healthcare, Impacting Patient Well- Being And Increasing Medical Costs. This Study Focuses On Predicting Hospital Readmission Rates For Diabetic Patients Using Machine Learning Techniques. By Analyzing Patient Demographics, Medical History, And Hospitalization Details, We Aim To Identify Key Risk Factors Contributing To Readmission. The Study Employs Various Classification Models, Including Logistic Regression, Decision Trees, Random Forests, And XGBoost, To Determine The Most Effective Predictive Approach. Our Findings Indicate That Certain Patient Attributes, Such As Time Spent In The Hospital, Number Of Inpatient Visits, And Medication Changes, Play A Significant Role In Readmission Likelihood.

Author: Ritesh SekarAVV | VikasB | AbhinandhanPS | Dr. S Nithya Roopa

Area of research: NA

Due To Rapid Increase In The Production And Consumption Processes, Societies Generate As Well As Reject Solid Materials Regularly From Various Sectors – Agricultural, Commercial, Domestic, Industrial And Institutional. The Considerable Volume Of Wastes Thus Generated And Rejected Is Called Solid Wastes. In This Study We Evaluate The Current Status And Identify The Major Problems. Total Solid Waste Generated In Tons/day Of Select Area Would Be Proportionate To The Population Of Specific Region In That Specific Year.To Develop Sustainable Management Of Solid Waste That Would Be Collected In Selected Area It Includes Manufacturing Of Paver Block ( Plastic Bottle Pieces, Steel Pieces, Coconut Shell, Metals) Vermicomposting (biodegradable Agriculture Waste , Food Waste , Domestic Kitchen Waste) And Suggestions Of Incinerator For Medical Waste , And Remaining Waste That Cannot Be Recyclable For Example Bubble Wrap, Plastic Bags, Ceramics , Household Glass Mirror Etc. That Dispose Through Landfilling In Road Construction. The Main Objective Of This Study Is To Maximum Use Of Waste To Get Sustainable Outcomes.

Author: Abhishek Jadhav | Aditya Pawar | Omkar Yerunkar | Prabhakar Shivatare

Area of research: CSE

Federated Learning (FL) Has Emerged As A Critical Paradigm For Collaborative Model Training In Privacy-constrained Domains, Particularly In Healthcare. This Study Presents A Comprehensive FedAvg-based Framework For Brain Tumor Detection From Magnetic Resonance Imaging (MRI) Scans, Employing Three Geographically Distributed Institutions As Local Clients And A Central Server For Global Aggregation. Each Client Trains An Identical Convolutional Neural Network (CNN) Model Using Institution-specific Subsets Of The BraTS 2020 Dataset, With Preprocessing Steps Including Skull Stripping, Intensity Normalization, And Uniform Resizing To 224×224 Pixels. Over 50 Communication Rounds, Local Models Perform Two Epochs Of Stochastic Gradient Descent Per Round, Contributing Data-weighted Parameter Updates To The Server. The Global Model, Initialized With Xavier Initialization, Converges Rapidly, Achieving A Validation Accuracy Of 96.2% By Round 30 And Stabilizing Between 95% And 97% By The Final Round. Comparative Analysis Against A Centralized Baseline—trained On Pooled Data—shows The Federated Framework Attains 96.5% Accuracy, Indicating Negligible Performance Degradation Despite Strict Privacy Constraints. Additional Evaluation Metrics Include Precision (95.8%), Recall (96.0%), And F1-score (95.9%), Demonstrating Balanced Classification Performance. Resource Utilization Metrics Reveal That Federated Training Incurs Only A 12% Increase In Training Time Relative To Centralized Training, Underscoring The Framework’s Efficiency. The Proposed Methodology Preserves Patient Privacy By Keeping Raw MRI Data Localized While Delivering Near-centralized Performance, Making It A Viable Solution For Multi-institutional Medical Imaging Collaborations. This Work Lays The Groundwork For Future Enhancements, Such As Integrating Secure Aggregation, Differential Privacy, And Personalized Model Fine-tuning, To Further Strengthen Privacy Guarantees And Model Personalization.

Author: Amruta Vijayakumar kavalapure | Anusha K N | Bhuvana S Kumar | Harshitha B | Mrs. Maria Rufina P

Area of research: CSE

Agoraphobia Is Frequently Overlooked Due To Low Mindfulness And Vacillation In Seeking Help. This Design Implements A Machine Literacy- Grounded System For Early Discovery Using Responses From A 10- Question Dataset. After Applying Mode Insinuation And Marker Garbling For Preprocessing, We Trained Several Bracket Models Including SVM, Decision Tree, Random Forest, Naive Bayes, And KNN. The Model With The Stylish Delicacy Was Named For Deployment. Druggies Can Interact With The System Through A Simple Interface That Accepts Quiz- Grounded Responses, Descriptive Textbook, And Voice Input( Under Development). In Addition To Prognostications, The Platform Offers Relaxation Tools Like Games And Links To Yoga And Contemplation Coffers, Making It Useful For Individualities And Internal Health Professionals Likewise.

Author: Neha Premnath D | Sadhana M S | Sanjana S | Anagha H R | Lavanya S

Area of research: CSE

Urban Traffic Congestion Has Become One Of The Major Issues In Modern Cities. As City Populations Grow And Vehicle Usage Increases, Existing Road Systems Struggle To Manage The Traffic Load Effectively. This Results In Long Delays, Energy Wastage, Increased Pollution, And Decreased Mobility. Traditional Traffic Control Methods Rely On Fixed-timing Signals That Follow Preset Schedules Without Considering Real-time Traffic Density. This Leads To Significant Inefficiencies And Commuter Frustration. To Tackle This Issue, We Propose An Intelligent Traffic Signal Control System That Dynamically Adjusts Signal Durations Based On Current Traffic Conditions. This System Integrates Internet Of Things (IoT) Devices With Image Processing Techniques. Specifically, Haar Cascade Classifiers Are Used To Detect Vehicles And Measure Traffic Density Efficiently. By Adapting Signal Timings According To Actual Traffic Flow At Intersections, The System Ensures Smoother Vehicle Movement And Minimizes Unnecessary Delays. The Use Of Decentralized, IoT-based Microcontrollers Enhances The System’s Flexibility And Reduces Reliance On Central Servers, Making It More Resilient In Practical Deployments. Simulated Results Demonstrate Notable Improvements In Reducing Traffic Delays, Optimizing Resource Use, And Enhancing Travel Experiences. This Adaptive Approach Paves The Way For Smarter Cities, Contributing To Reduced Environmental Impact And Improved Urban Quality Of Life.

Author: Rishika S | Shravya S | Siri N | Sushna Subramanya K | Harshitha B

Area of research: Computer Science And Engineering

It's Been Quite Difficult To Teach Students Over An Online Platform And Get The Lesson Interesting Amidst The COVID-19 Pandemic. Due To This Reason, There Was A Necessity Of A Dustfree Classroom For Kids. This Article Uses MediaPipe And OpenCV To Offer An Interesting Paint Application Which Recognizes Hand Gestures And Traces Hand Joints. The Application Makes Use Of Hand Gestures To Provide Users With A User-friendly Approach To Human Computer Interaction (HCI). HCI's Overall Objective Is To Enhance Human-computer Interaction.

Author: Nisha P | Pavana Rao | Sinchana C R | Souparna D S | Shyleshwari M Shetty

Area of research: ECE

Reversible Logic Gates Are Critical Components In The Field Of Quantum Computing And Low-power Digital Circuits, As They Allow For The Retrieval Of Input States From Output States Without Any Loss Of Information. This Property Ensures Minimal Energy Dissipation, Aligning With The Principles Of Reversible Computation. Their Applications Extend To Quantum Computing, Cryptographic Systems, And Error Detection And Correction Mechanisms. This Paper Presents A Novel Design For A Reversible D Flip-Flop (RDFF) And A 4-bit Linear Feedback Shift Register (LFSR). The Linear Feedback Shift Register (LFSR) Is Built Utilizing Four Register-Driven D Flip-Flops (RDFFs) And A Feedback Mechanism That Incorporates A Feynman Gate. This Configuration Successfully Showcases The Capacity To Generate Pseudo-random Sequences. The LFSR Design Demonstrates A 10%improvement In Total Reversible Logic Implementation Cost And27% Enhancement In Quantum Cost, Making It A More Resource Efficient Option For Reversible Computing. This Work Makes A Valuable Contribution To The Field Of Reversible Computing By Offering Efficient Designs For Essential Components.

Author: J.Vanaja | K.Suresh Kumar

Area of research: CSE

The Jarvis Application Is A Voice-activated Virtual Assistant Designed To Automate Daily Tasks Through Simple Voice Commands. It Offers A Range Of Functionalities, Including Personalized Greetings, Web Searches (Google, YouTube, And Wikipedia), Real-time Weather And Time Updates, And Alarm Setting. Users Can Also Control Applications, Manage Schedules, Perform Calculations, And Access News Updates. Jarvis Supports Messaging Tasks On WhatsApp And Includes Password Protection, A “Remember” Function For Storing Information, And A Focus Mode To Minimize Distractions. Additionally, It Features Fun Elements Like Rock Paper Scissors And Live IPL Score Tracking. The Application Is Packaged As An Executable File, Making It Compatible Across Different Systems.The Jarvis System Provides A Practical And Interactive Assistant, Enhancing Productivity While Offering A User-friendly, Hands-free Experience. Overall, Jarvis Combines Advanced Technology With Practicality, Delivering A Highly Interactive, Hands-free Experience That Transforms The Way Users Manage Their Daily Lives.

Author: Disha H N | Disha H N | Kruthika H R | Spoorthi N K | Ashwini M S

Area of research: Computer Science And Engineering

The Smart Canteen System Is An Advanced Solution Designed To Digitize And Optimize The Operations Of Traditional Canteens By Integrating Modern Technologies. It Enhances Customer Convenience And Staff Efficiency Through Features Such As QR Code Scanning, Real-time Menu Updates, Automated Billing, And Cashless Payments. Customers Can Place Orders Via A Mobile App Or Kiosk Interface, Thereby Eliminating Long Queues And Manual Intervention. Real-time Order Tracking And Notifications Further Improve The User Experience By Ensuring Transparency And Reducing Wait Times. From The Administrative Perspective, The System Provides Tools For Inventory Tracking, Sales Monitoring, And Data-driven Decision-making Through Analytical Reports. By Integrating IoT And Data Analytics, It Predicts Demand Patterns, Reduces Food Wastage, And Supports Cost-effective Management. The Smart Canteen System Is A Scalable And Sustainable Solution Suitable For Educational Institutions, Offices, And Similar Environments, Reflecting The Transformative Power Of Technology In Modernizing Food Services. Motivation- The Smart Canteen System Is Driven By The Increasing Demand For Efficiency, Convenience, And Accuracy In Traditional Canteen Operations. Conventional Systems Often Suffer From Long Queues, Manual Order Processing, And Cash- Based Transactions, Which Lead To Time Wastage, Human Error, And Customer Dissatisfaction. With The Growing Reliance On Technology In Daily Life, There Is A Clear Need For A Digital Solution That Streamlines These Processes. This Project Aims To Eliminate Delays, Reduce Errors, And Provide Real-time Updates Through Features Like Cashless Payments, Automated Billing, And AI-powered Assistance. It Not Only Saves Time For Both Customers And Staff But Also Improves Inventory Management, Reduces Food Wastage, And Enhances Overall Service Quality. By Aligning With Modern User Expectations And The Broader Trend Of Digital Transformation, The Smart Canteen System Offers A Scalable And Sustainable Approach To Modernizing Institutional Food Services [7].

Author: Anwitha | Dhanushree BA | Monika CV | Darshini MS

Area of research: CSE

Concern Over Insurance Fraud In The Machine Sedulity Is Growing, As It Can Affect In Significant Financial Losses And Advanced Decorations For Law- Abiding Policyholders. This Study Offers A Machine Knowledge- Predicated System For Further Directly Relating False Claims. To Address Class Imbalance, We Use A Kaggle Dataset And The SMOTE Fashion. With A Python Flask- Predicated Web Interface That Lets Stoners Enter Claim Details And Get Immediate Fraud Discovery Results, Our System Predicts Fraud Using Random Forest, Decision Tree, And Logistic Regression Models.

Author: Pushpavalli P | Shamitha M | Aishwarya S | Shyleshwari M Shetty

Area of research: Biomedical Engineering

The Proposed System Is A Deep Learning-based Dental Image Classification Tool That Uses Deep Metric Learning Techniques For Accuratedental Conditions Identification, Such As Cavities, Implants, Fillings, Implanted Teeth, Impacted Teeth, And Root Canals Play A Critical Role In The Early Diagnosis And Treatment Planning In Dentistry. This Project Proposes An Automated, Deep Learning-based Solution For Teeth Classification Using Image Processing And Deep Metric Learning Techniques. Leveraging DenseNet121, The System Is Trained To Detect And Classify Various Dental Conditions Identification With High Accuracy. The Classification Is Achieved Through A Well-defined Pipeline Involving Image Preprocessing, Segmentation, Image Splitting, And Feature Extraction, Enabling The Model To Handle Variability In Image Quality And Tooth Structure.The System Is Deployed As A User-friendly Web Application Built Using Streamlit, Allowing Users To Register, Log In, And Upload Dental Images For Instant Classification Results. The Application Processes The Image, Classifies The Dental Condition, And Displays Performance Metrics Such As Accuracy, Confusion Matrix, And ROC Curve. This Solution Aims To Assist Dental Professionals, Researchers, And, In Forensic Cases, By Providing An Accessible Tool For Image-based Dental Condition Identification, Thus Enhancing Clinical Decision-making Through Automated And Efficient Image Analysis.

Author: Vidhya.S | Pritika.R.G | Rashmi Ifrashiya.A | Soundarya.S

Area of research: AIML

Bankruptcy Prediction Is A Critical Area In Financial Risk Assessment, Supporting Timely Decisions For Investors, Regulators, And Institutions. This Study Presents A Comparative Analysis Of Multiple Machine Learning Models, Including Traditional Algorithms (Decision Tree, Naive Bayes), Deep Learning Methods (CNN, LSTM), And Hybrid Approaches (XGBoost + ANN, Decision Tree + Gaussian), Applied To An Imbalanced Financial Dataset From Polish Companies. The Dataset Poses Real-world Challenges Such As Class Imbalance And Feature Noise, Which Are Addressed Through Data Preprocessing, Feature Selection, And Resampling Techniques. The Proposed Hybrid Models Integrate The Strengths Of Ensemble Learning And Neural Networks, Improving Classification Performance On Minority (bankrupt) Classes. Evaluation Using Metrics Like Precision, Recall, And F1-score Demonstrates That Hybrid And Deep Learning Models Outperform Traditional Classifiers, With The XGBoost–ANN Model Achieving The Best Overall Results. Feature Importance Analysis Further Reveals The Most Influential Financial Indicators Contributing To Bankruptcy Prediction. This Work Offers A Robust, Adaptable Framework For Handling Imbalanced Datasets In Financial Domains, Contributing Practical Insights For Early Risk Detection And Decision-making.

Author: Aishwarya M | Anu Shree R M | Dhanyashree T N | Meghana Shree S | Dr.Vishwesh J

Area of research: CSE

Heart Disease Is One Of The Leading Causes Of Mortality Worldwide, Necessitating Early And Accurate Detection For Effective Treatment And Prevention. This Project Focuses On Developing A Predictive Model For Heart Disease Detection Using Logistic Regression, A Robust Statistical Method Widely Used For Binary Classification Problems. The Primary Goal Is To Analyze Patient Data And Predict The Likelihood Of Heart Disease Based On Various Clinical And Demographic Attributes.

Author: Aishwarya R | Chitra M P | Hema L Patel | M Ishani Kuttappa | Dr. Madhu M Nayak

Area of research: CSE

Fake Product Identification Is A Critical Process In Combating Counterfeit Goods, Ensuring Consumer Safety, And Protecting Brand Integrity. One Of The Biggest Challenges In Today's Retail Market Is The Counterfeiting Of Products. Counterfeiting Products Are Just Low-quality Copies Of Some Genuine Brand. Many Different Methods Have Been Adopted From Time To Time To Combat The Counterfeiting Of The Products Such As RFID Tags, Artificial Intelligence, Machine Learning, QR Code-base System, And Many More.Counterfeit Products, Often Designed To Mimic Genuine Items, Can Pose Significant Risks, From Financial Loss To Health Hazards. This Growing Problem Necessitates The Development Of Robust Methods And Technologies To Identify Fake Products Quickly And Accurately To Address This Challenge, Various Methodologies Have Been Adopted, Including Physical Inspection, Digital Authentication Systems, And Machine Learning Algorithms. Physical Inspection Involves Analyzing Packaging, Materials, And Labels For Inconsistencies. Digital Tools, Such As QR Codes And Serial Number Verification, Provide Real-time Authentication. Machine Learning And AI Enhance The Process By Analyzing Patterns And Detecting Anomalies In Product Features With High Precision. These Approaches Are Implemented In Collaboration With Manufacturers, Retailers, And Consumers To Create A Seamless Verification Process.

Author: Amrutha S L | Chinmaye Patel N K | Dhaarini Lokesh | Harshitha P P | Shreelakshmi C M

Area of research: CSE

The "Currency Detection System" Is An Innovative Application Designed To Distinguish Between Genuine And Counterfeit Currency Notes. Leveraging Advanced Deep Learning Models, Feature Matching Techniques, And An Intuitive User Interface, The System Provides An Efficient And User-friendly Approach To Detecting Fake Currency. With Applications In Retail, Banking, And Public Sectors, This Project Aims To Reduce The Prevalence Of Counterfeit Notes And Promote Financial Security.

Author: Manjula M | Bhagyashree Peddi | Dhanalakshmi G | Dr. Madhu M Nayak

Area of research: Computer Science And Engineering

An Innovative Attendance System Utilizing Face Detection Technology Is Presented, Aimed At Improving The Efficiency And Accuracy Of Attendance Tracking. This System Integrates Computer Vision With Advanced Deep Learning Techniques, Enabling Reliable Recognition Of Individuals And Real-time Attendance Logging. Convolutional Neural Networks (CNNs) Are Employed For Face Detection And Recognition, Establishing A Robust Alternative To Traditional Attendance Methods. With High Detection Accuracy, Rapid Processing Times, And Comprehensive Data Security Protocols, This System Is Well-suited For Implementation In Educational Institutions, Corporate Environments, And Secure Access Management. Experimental Results Indicate A Detection Accuracy Of 98.6% And An Average Verification Time Of Under 1.5 Seconds, Underscoring The Effectiveness Of Face Recognition Technology In Automated Attendance Systems.

Author: Thinesh T | Mr. S. Chandrasekar | Sivasakthi S | Sivanthamil M | Suman Raj R

Area of research: Machine Learning

This Research Presents An Automated Approach To Analyzing Brain States During Anesthesia Using Convolutional Neural Networks (CNNs) And Long Short-Term Memory (LSTM) Networks. By Leveraging The Spatial Feature Extraction Power Of CNNs And The Temporal Sequence Processing Capabilities Of LSTMs, The Model Effectively Classifies Brain States From EEG Signals. The System Identifies Key States Such As Consciousness, Light Anesthesia, Deep Anesthesia, And Emergence. Extensive Experiments On EEG Datasets Show That The Proposed CNN-LSTM Hybrid Architecture Outperforms Traditional Machine Learning Methods In Accuracy. This Method Offers Real-time, Objective, And Precise Monitoring Of Brain States, Aiding Anesthesiologists In Clinical Decision-making. The Research Paves The Way For Safer Anesthesia Practices By Integrating Advanced Deep Learning Technologies For Reliable Brain State Classification.

Author: Ankitha D D | Harshitha M G | Manjula C | Meghana V Mathad | Rummana Firdaus

Area of research: CSE

As Healthcare Increasingly Embraces Cloud-based Technologies, It Faces Critical Issues Such As Transmission Delays And Heightened Security Risks Due To Centralized Data Storage. Traditional Cloud Setups Often Fall Short In Delivering Real-time Patient Data And Safeguarding Sensitive Information From Unauthorized Access. To Overcome These Challenges, This Project Presents A Secure Authentication-Based Patient Report Transmission System Powered By Edge Computing. By Handling And Verifying Data At Local Edge Nodes—closer To Where It Originates—the System Significantly Reduces Latency And Enables Prompt Data Delivery, Which Is Essential For Time-sensitive Medical Decisions. This Decentralized Model Not Only Improves Processing Speed But Also Enhances The Overall Efficiency Of Data Management. To Ensure Robust Security, The System Incorporates Multi-factor Authentication (MFA) And Role-based Access Control (RBAC), Which Together Help Confirm User Identities And Limit Access Based On Professional Roles. This Innovative Framework Offers A Reliable, Secure, And Fast Method For Transmitting Healthcare Data, Making It Highly Effective For Critical Care Environments That Demand Both Speed And Privacy.

Author: Mrs. K. Ramya | Manojalex G | Mohamed Irfan J | Kathirvalavan G | Prabu A

Area of research: Computer Science And Engineering

This Project Introduces A Smart, Real-time Health Monitoring System Built Using Internet Of Things (IoT) Technology To Help Respond Quickly To Cardiac Emergencies. It Uses A Small Sensor Called The MAX30100 To Track A Person’s Heart Rate And Oxygen Levels. These Readings Are Processed By A Compact Microcontroller (NodeMCU ESP8266), Which Checks For Any Signs Of Abnormal Heart Activity Like Unusually Fast (tachycardia) Or Slow (bradycardia) Heartbeats. If Something Unusual Is Detected, The System Automatically Sends An Emergency Text Message Via A GSM Module. This Message Includes The Person's Location, Thanks To A Built-in GPS Module (NEO-6M). In Addition, Caregivers Or Family Members Can View The Patient’s Live Health Data Remotely Using The Blynk App On Their Smartphones. The System Is Designed To Be Portable, Battery-operated, And Reliable Even In Remote Areas With Limited Medical Access. Overall, It Helps Speed Up Emergency Responses And Could Be A Valuable Tool For Improving Heart Health Monitoring On A Larger Scale.

Author: Sahana K | Sakshi Annasab Kamate | Srashti Kumbar | Yellamelli Sahithi | Dr. Raviraj P

Area of research: Management Studies

The Study Entitled “A STUDY ON THE EFFECTIVENESS OF TRAINING AND DEVELOPMENT AMONG EMPLOYEES IN BHARATH RUBBER INDIA LIMITED, MADURAI This Study Investigates The Effectiveness Of Training And Development Programs Among Employees At Bharath Rubber India Limited, Madurai. Utilizing Primary Data Collected Through Structured Questionnaires, The Research Aims To Assess How These Programs Influence Employee Performance, Motivation, And Job Satisfaction. The Study Is Structured Into Five Chapters: The First Introduces The Concept, Need, And Scope Of Training And Development, Along With A Literature Review; The Second Provides A Profile Of Bharath Rubber India Limited; The Third Outlines The Research Design, Including Objectives, Limitations, And Methodology; The Fourth Presents Data Analysis Using Tools Such As Simple Percentage Analysis, Correlation, And Regression; And The Fifth Discusses Findings, Offers Suggestions, And Concludes The Study. Key Findings Suggest That While A Majority Of Employees Acknowledge The Importance Of Training In Enhancing Knowledge And Skills, There Is Room For Improvement In Areas Such As The Quality Of External Training Agencies And The Incorporation Of Modern Training Methods. The Study Concludes That Effective Training And Development Are Crucial For Organizational Growth And Recommends Regular Feedback Mechanisms And The Adoption Of Contemporary Training Techniques To Further Enhance Employee Development.

Author: Jayaharini B S | Mr.Imayavan B

Area of research: CSE

Women's Safety Is Becoming An Increasingly Pressing Topic In India And Other Nations. The Fundamental Difficulty With The Police Handling Of These Incidents Is That They Are Limited In Their Ability To Respond Swiftly To Distress Calls. These Limits Include Not Knowing The Location Of The Crime And Not Knowing The Crime Is Occurring At All At The Victim's End, Making Reaching The Police Confidently And Discreetly Difficult. To Avoid These Circumstances, This Project Develops A Mobile Application That Provides Women With A Dependable Option To Make An Emergency Call, Send A Message, And Update Her Whereabouts To The Police As Well As Her Family's Close Relatives.

Author: Acharya Meghpriya Ramesh | C Poojitha | Deekshitha H | Likitha S V | Nischitha B S

Area of research: Managerial Economics

Fluctuations In Global Oil Prices Have Emerged As A Key Factor Influencing Inflation Trends Across The World, With India Being Particularly Vulnerable Due To Its Heavy Reliance On Crude Oil Imports. As Oil Prices Surge, The Indian Economy Faces Mounting Pressure From Increased Transportation And Production Expenses. These Rising Costs Contribute To An Overall Escalation In The Prices Of Goods And Services, Diminishing Consumer Purchasing Power. This Paper Investigates The Inflationary Effects Of Oil Price Hikes In India, Emphasizing Their Influence On Everyday Living Expenses And Consumption Habits. The Research Also Reviews How Reduced Disposable Income Affects Spending Patterns And Evaluates The Government’s Interventions Aimed At Countering These Economic Pressures. Findings Suggest A Notable Inverse Relationship Between Rising Oil Prices And Consumer Spending, Indicating A Significant Strain On Domestic Demand. The Study Highlights The Necessity For Strategic Policy Initiatives To Shield Consumers And Maintain Economic Equilibrium Amid Ongoing Global Oil Price Volatility.

Author: Santhiya S | Dr. M. D. Chinnu, Assistant Professor

Area of research: Managerial Economics

Capital Budgeting Stands As A Crucial Element Within Financial Decision-making, Especially For Large-scale Industries, As It Empowers Firms To Undertake Strategic Investment Choices Aligned With Their Long-term Corporate Objectives. This Research Explores The Real-world Application Of Various Capital Budgeting Techniques, Such As Net Present Value (NPV), Internal Rate Of Return (IRR), Payback Period, Discounted Payback Period, Profitability Index, And Modified Internal Rate Of Return (MIRR). The Effectiveness Of These Methods In Aiding Industries To Evaluate Investment Opportunities And Assess Associated Risks Is Closely Examined. Although Many Corporations Implement Advanced Discounted Cash Flow (DCF) Models, Practical Business Conditions Often Necessitate A Combination Of Both Traditional And Modern Methods. Elements Like Organizational Size, Industry Sector, Capital Framework, And Leadership Approach Significantly Influence The Choice Of Evaluation Techniques. Additionally, Contemporary Risk Assessment Tools, Including Real Options Analysis And Scenario Planning, Are Becoming More Common In Investment Decision-making Processes. Simpler Approaches, Such As The Payback Period Method, Continue To Hold Favor In Certain Sectors Due To Their Simplicity And Quick Results, Despite Known Limitations. This Paper Sheds Light On The Evolving Trends In Budgeting Strategies Across Various Industries And Regions, Underlining The Importance Of Matching Financial Decision-making Tools With Objectives For Sustainable Growth And Adaptability In A Constantly Changing Business Landscape.

Author: Sridharshana.S.U | Dr. M. D. Chinnu, Assistant Professor

Area of research: CSE

Artificial Intelligence (AI) Is Transforming A Wide Range Of Industries By Enabling Machines To Perform Tasks That Traditionally Required Human Intelligence, Such As Perception, Decision-making, And Natural Interaction. In The Context Of Recruitment, Traditional Interview Methods Often Focus Primarily On Technical Skills, Neglecting Important Aspects Like Emotional Intelligence And Candidate Confidence. This Paper Presents An AI-powered Mock Interview Evaluator That Offers A More Holistic Assessment By Analyzing Emotional Expressions And Confidence Levels In Real Time. The System Combines Convolutional Neural Networks (CNNs) For Facial Emotion Recognition And Recurrent Neural Networks (RNNs) For Analyzing Speech And Body Language. Trained On A Diverse Dataset Of Mock Interviews, The Model Can Detect Emotions Such As Happiness, Sadness, Anger, And Surprise, While Also Estimating Confidence Through Multimodal Analysis.

Author: Manasvi H | Deeksha V | M Sanjana | Anupama C Swamy | Rummana Firdaus

Area of research: Mechanical Engineering

The "Smart Movable Vehicles With Robotic Arm" Aims To Design And Develop A Semi-autonomous Mobile Robotic System Capable Of Performing Remote Operations In Dynamic Environments. This Vehicle Integrates Mobility With A Versatile Robotic Arm, Enabling It To Carry Out Tasks Such As Material Handling, Object Detection, And Pick- And-place Operations Efficiently. The System Is Designed To Operate Under Both Wired And Wireless Control Modes, Enhancing Its Adaptability And User Flexibility In Various Operational Scenarios. The Core Objective Of This Project Is To Simulate Real-world Industrial Or Rescue Applications Where Human Intervention Might Be Difficult Or Risky. The Vehicle’s Movement And Robotic Arm Actions Can Be Monitored And Controlled Through A Dual-interface System: One Through Direct Wired Commands And Another Through Wireless Modules Such As Bluetooth Or Wi-Fi. This Dual-control Approach Ensures Consistent Operation Even In Cases Where One Mode Fails Or Becomes Less Effective Due To Environmental Limitations. The Robotic Arm Is Actuated Using Servo Or DC Motors, Which Are Precisely Controlled To Execute Complex Tasks Like Gripping, Lifting, And Placing Objects. The Smart Vehicle Base Is Equipped With Motor Drivers And Microcontrollers (such As Arduino Or Raspberry Pi) That Interpret User Inputs And Manage Both Navigation And Arm Coordination. Sensors May Be Integrated To Enhance Environmental Awareness, Such As Obstacle Detection Or Camera-based Monitoring.

Author: Madhavan V | Ajith S | Karthikeyan S | Shanmuganathan P | Vinoth Kumar V

Area of research: Computer Science And Engineering

The Smart Traffic Monitoring And Violation Detection System Is An Advanced AI-powered Solution Designed To Revolutionize Traffic Management By Ensuring Real-time Monitoring, Automated Enforcement, And Enhanced Road Safety. Leveraging Computer Vision, Deep Learning, And IoT-based Surveillance, The System Intelligently Detects Violations Such As Helmetless Riding, Signal Jumping, With Exceptional Accuracy. High-resolution Cameras Continuously Capture Live Traffic Feeds, While OCR Technology Extracts Vehicle Number Plate Details For Instant Offender Identification. With YOLOv11 For Object Detection And Convolutional Neural Networks (CNNs) For Number Plate Recognition, The System Achieves Precise And Efficient Violation Detection. Upon Detection, Automated SMS Alerts Are Dispatched To Violators, Providing Fine Details And Enforcement Actions. Integrated With RTO Databases, It Tracks Repeat Offenders, Issuing Escalating Penalties, Including Potential Registration Cancellation For Persistent Violations. By Eliminating Manual Intervention, This System Optimizes Traffic Law Enforcement, Reduces Human Workload, And Ensures Seamless, Technology-driven Compliance. The Fusion Of Real-time AI Processing, Automated Data Retrieval, And Instant Penalty Notification Transforms Traffic Monitoring Into A Smart, Proactive, And Efficient System, Significantly Improving Urban Mobility And Road Discipline While Minimizing Accidents And Fatalities.

Author: Shapna Rani E | Eraiarul K | Karthick M | Darwin Shiyam B | Hariharan T

Area of research: CSE

Individuals Frequently Disregard How Important It Is To Wear Helmets, Which Is Tragic. A Helmet Reduces Your Risk Of Getting A Serious Brain Injury And Dying By Deflecting Most Of The Impact Energy That Would Otherwise Hit Your Head And Brain During A Tumble Or Collisions. In India, It Is Against The Law To Operate A Motorbike Or Scooter Without A Helmet, Which Has Increased Fatalities As Well As Crashes. The Existing System Mostly Relies On Surveillance Footage For Keeping Up With Traffic Violations, Necessitating A Close - Up Of The License Plate By Traffic Police In The Case That The Motorcyclist Lacks A Helmet. Yet, This Necessitates A Substantial Amount Of Personnel And Time Considering The High Frequency Of Traffic Violations And The Rising Everyday Use Of Motorcycles. Imagine If There Was An Algorithm That Monitored Traffic Infractions, Such As Driving A Motorbike With No A Helmet, And, If Any Were Identified, Generate The License Plate Of The Vehicle That Committed The Violation. Helmet And License Plate Detection Using A Neural Network Is Proposed In This Paper. There Will Be Two Phases. Initially, We Check To See If The Riders Are Wearing Helmets. If Not, A Second Step Is Used To Find Their License Plate. To Identify Unauthorized Vehicles, We Also Look For License Plates On Passing Vehicles.

Author: Prabhu P | Ragupriyan P | Sathish Kumar S | Sridesinguraja S | Rahul christober. F

Area of research: CSE

Pneumonia Is A Severe Lung Infection That Demands Timelyandprecise Diagnosis.This Project Utilizes Convolutional Neural Networks (CNNs) To Detect And Classify Pneumonia From Chest X-ray Images. A Large Dataset Is Used To Train The Model, Ensuring High Accuracy. Image Preprocessing And Augmentation Enhance Detection Performance. The System Effectively Differentiates Between Normal And Infected Lungs. Evaluation Metrics Like Accuracy And F1-score Confirm The Model’s Reliability. This Solution Supports Rapid, Scalable, And Affordable Diagnosis In Medical Settings.

Author: Amrutha K | Anusha S T | Kousar K | Maanasa Nagaraju | Asha Rani M

Area of research: Economics

Dynamic Pricing Has Become A Cornerstone Strategy In The Digital Marketplace, Allowing Businesses To Adjust Prices In Real Time Based On Factors Like Demand, Competition, And Consumer Behavior. This Approach Leverages Advanced Data Analytics And Machine Learning Algorithms To Optimize Pricing, Aiming To Maximize Revenue And Enhance Market Competitiveness. Industries Such As E-commerce, Hospitality, And Transportation Have Widely Adopted Dynamic Pricing Modelsin Order To React Quickly To Changes In The Market. However, The Effectiveness Of These Strategies Varies Across Sectors And Depends On Factors Like Consumer Acceptance And Technological Infrastructure. This Study Aims To Analyze The Effectiveness Of Dynamic Pricing Strategies In The Digital Era By Examining Their Impact On Consumer Behavior, Revenue Optimization, And Market Dynamics. It Will Explore The Benefits And Challenges Associated With Implementing Dynamic Pricing, Considering Ethical Considerations And Consumer Perceptions. Through A Comprehensive Literature Review And Case Studies, The Research Seeks To Provide Insights Into Best Practices For Businesses Looking To Implement Dynamic Pricing Strategies Effectively. The Findings Will Contribute To A Deeper Understanding Of How Dynamic Pricing Can Be Leveraged To Achieve Business Objectives While Maintaining Customer Trust And Satisfaction.

Author: Sathveka D K | Dr.M D Chinnu

Area of research: Structural Engineering

Soil Structure Interaction (SSI) Is The Response Of Soil That Influence The Motion Of The Structure. Soil Structure Interaction Is Prominent For Heavy Structure, Especially For High Rise Building Located On Soft Soil. Incorporation Of Soil Interaction Effect Will Reduce The Base Shear And Flexibility Of Soil. Because Of This The Stiffness Of The Building Is Getting Reduced Resulting, Increase In The Natural Period Of The Structure During Earthquake. One Cause Of These Deviations Is Base-slab Averaging, In Which Spatially Variable Ground Motions Within The Building Envelope Are Averaged Within The Foundation Footprint Due To The Stiffness And Strength Of The Foundation System. Another Cause Of Deviation Is Embedment Effects, In Which Foundation-level Motions Are Reduced As A Result Of Ground Motion Reduction With Depth Below The Free Surface. Interaction Of Pile Foundation With Wave Propagation Below The Base Slab, Which Can Further Modify Foundation-level Motions At The Base Of A Structure.

Author: Dipak H. Vaidya | Mr. Girish Savai

Area of research: Commerce

The Project Work Is Entitled A " A STUDY ON CUSTOMER PREFERENCES FOR COSMETIC BRANDS WITH REFERENCE TO COIMBATORE CITY" With Special Reference To The Cosmetic Brands. The Primary Objective Of This Study Is To Measure And Analyse The Perceptions And Attitude Of The Public For Cosmetic Brands. The Main Objective Is To Compare The Cosmetic Brands Between Customers.

Author: Dr.K.S.Nirmal Kumar | Monisha Violet. J

Area of research: CSE

Breast Cancer Remains A Significant Global Health Concern, Impacting Millions Of Women Each Year. Timely Detection And Precise Diagnosis Are Essential To Enhancing Treatment Success And Lowering Death Rates. Histopathological Imaging Is Widely Utilized For Diagnosing Breast Cancer, But Interpreting These Images Accurately Often Requires Specialized Medical Expertise, Which May Not Be Readily Available In All Clinical Environments. The Dataset Used In This Study Comprises Breast Tissue Images Labeled To Reflect The Presence Or Absence Of Cancer. A Convolutional Neural Network (CNN) Was Employed To Automatically Extract Meaningful Features From The Images, Followed By A Fully Connected Layer To Perform Classification. The Model Was Optimized By Minimizing Prediction Error Using A Suitable Loss Function And Optimization Technique. To Assess Its Effectiveness, The Model's Performance Was Measured Using Metrics Such As Accuracy.

Author: Mrs.T.Geetha | Lakshmi R | karthiga S | Lashiya M | Preethi B