Sheela S. Maharajpet
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Sheela S. Maharajpet has a total of 27 co-authors. Here are the top co-authors by number of shared publications:
- Abhilash H P
Assistant Professor, School of CSA, REVA University, India
3 shared publications - Manish Kumar Thakur
Associate Professor, Department Of MCA, Acharya Institute of Technology, Bangalore
2 shared publications - Mithun G P
Acharya Institute of Technology,Dept. of MCA,Bangalore,India,560107
2 shared publications Ananya V Hegde
Department of MCA, Acharya Institute of Technology, Bangalore, India
1 shared publication- Anjali Rao
Final Year Student Dept. Of MCA Acharya Institute of Technology Bangalore
1 shared publication - Chiranjeevi S
Acharya Institute of Technology, Bangalore -560107, Karnataka, India
1 shared publication
Latest Publications
- 6 October 2026
- Published by Springer Nature
- p. 521-534
Development of a Machine Learning Model for Early Prediction of Hydronephrosis
- 6 October 2026
- Published by Springer Nature
- p. 521-534
Development of a Machine Learning Model for Early Prediction of Hydronephrosis
- 26 September 2025
- Published by Inventive Research Organization in Recent Research Reviews Journal
- Vol. 4 (2) , 264-280
Intelligent Real-Time Crowd Density Estimation for Proactive Event Safety: A Machine Learning Approach
High-density crowd events like public concerts, sporting events, or religious festivals represent significant safety challenges due to high crowd density. Methods of monitoring crowds using manual observation or passive surveillance often don't provide real-time information. Therefore, we present a real-time crowd density estimation solution that utilizes YOLOv8 for people detection and CSRNet for density estimation. The crowd density estimation system utilizes live video feeds from direct surveillance cameras/CCTV or drone footage. The system will assess the crowd density in four levels across four distinct areas: Low, Medium, High, or Critical crowd density. The density estimator has a web-based dashboard that provides real-time analytics reports, heatmap density estimates, and historical records, which can assist in making quick, informed decisions and assessments following an event. The system is validated on benchmark datasets and real-world video streams with 95.3% detection accuracy, 7.4 MAE in crowd counting and 28 FPS processing with off-the-shelf GPU hardware. The results show high accuracy with low latency, making it feasible for real-world applications for large-scale events. The main contributions of the work include using YOLOv8 integrated with CSRNet to jointly detect and estimate crowds, developing a real-time dashboard to provide transparent crowd analytics, and system validation with quantitative metrics and real-world evidence.
- 26 September 2025
- Published by Inventive Research Organization in Recent Research Reviews Journal
- Vol. 4 (2) , 264-280
Intelligent Real-Time Crowd Density Estimation for Proactive Event Safety: A Machine Learning Approach
High-density crowd events like public concerts, sporting events, or religious festivals represent significant safety challenges due to high crowd density. Methods of monitoring crowds using manual observation or passive surveillance often don't provide real-time information. Therefore, we present a real-time crowd density estimation solution that utilizes YOLOv8 for people detection and CSRNet for density estimation. The crowd density estimation system utilizes live video feeds from direct surveillance cameras/CCTV or drone footage. The system will assess the crowd density in four levels across four distinct areas: Low, Medium, High, or Critical crowd density. The density estimator has a web-based dashboard that provides real-time analytics reports, heatmap density estimates, and historical records, which can assist in making quick, informed decisions and assessments following an event. The system is validated on benchmark datasets and real-world video streams with 95.3% detection accuracy, 7.4 MAE in crowd counting and 28 FPS processing with off-the-shelf GPU hardware. The results show high accuracy with low latency, making it feasible for real-world applications for large-scale events. The main contributions of the work include using YOLOv8 integrated with CSRNet to jointly detect and estimate crowds, developing a real-time dashboard to provide transparent crowd analytics, and system validation with quantitative metrics and real-world evidence.
- 18 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 171-177
Hybrid Face Recognition Model for Student Attendance Tracking Using Deep Learning
Manual attendance registers in educational systems is inefficient and prone to human error; therefore, they need an automated solution that is efficient and accurate. This paper describes a hybrid face recognition system combining FaceNets deep convolution neural network with triplet loss, which extracts 128-dimensional face embeddings. The attendance monitoring system has a dual input capability of group photo uploads where 2-3 images are combined into a single file to improve face detection accuracy, and live attendance captured via specific camera utilization in a web application. The face detection part of the program uses Open CVs Haar Cascade and the attendance is stored as embedding's in a SQLite database for fast face matching. The report offers an online dashboard with real time attendance reporting, historical attendance records, and CSV file download capabilities. Implemented on a dataset of 30 images of students followed by (5) live feeds, the system achieved 98.5% accuracy which we believe is better than current manual and automatic methods date-to-date (HOG+PCA of 85.7%). The proposed method has improved scalability, improves potential security, and provides ease of administration in educational settings.
- 18 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 171-177
Hybrid Face Recognition Model for Student Attendance Tracking Using Deep Learning
Manual attendance registers in educational systems is inefficient and prone to human error; therefore, they need an automated solution that is efficient and accurate. This paper describes a hybrid face recognition system combining FaceNets deep convolution neural network with triplet loss, which extracts 128-dimensional face embeddings. The attendance monitoring system has a dual input capability of group photo uploads where 2-3 images are combined into a single file to improve face detection accuracy, and live attendance captured via specific camera utilization in a web application. The face detection part of the program uses Open CVs Haar Cascade and the attendance is stored as embedding's in a SQLite database for fast face matching. The report offers an online dashboard with real time attendance reporting, historical attendance records, and CSV file download capabilities. Implemented on a dataset of 30 images of students followed by (5) live feeds, the system achieved 98.5% accuracy which we believe is better than current manual and automatic methods date-to-date (HOG+PCA of 85.7%). The proposed method has improved scalability, improves potential security, and provides ease of administration in educational settings.
- Open Access
- 1 July 2025
- Published by AMO Publisher in Scientia. Technology, Science and Society
- Vol. 2 (7) , 51-61
Hybrid Face Recognition Model for Student Attendance Tracking Using Machine Learning Algorithms
Attendance tracking in educational institutions is critical, the manual methods are very difficult, in particularly for large numbers of student populations. In this research novelty new methodology proposed work a face recognition based attendance monitoring system that employs deep learning prediction system and computer vision. This system aims to streamline processes, reduce fraud, and improve accuracy. This study has used robust face detection algorithms, combined with Histograms of Oriented Gradients (HOG) feature extraction to yield a comprehensive database of authorized students face detection using Deep Learning (DL), when integrated with Principal Component Analysis (PCA), Support Vector Machine (SVM), KNearest Neighbor (KNN), and CNN classification, improves both system performance and accuracy. Enrollment generates unique identifiers, and regular updates to a centric dataset in order to make attendance tracking easier. This study promises to regulate attendance in schools in a practical and precise way.
- 1 July 2025
- Published by AMO Publisher in Scientia. Technology, Science and Society
- Vol. 2 (7) , 51-61
Open AccessHybrid Face Recognition Model for Student Attendance Tracking Using Machine Learning Algorithms
Attendance tracking in educational institutions is critical, the manual methods are very difficult, in particularly for large numbers of student populations. In this research novelty new methodology proposed work a face recognition based attendance monitoring system that employs deep learning prediction system and computer vision. This system aims to streamline processes, reduce fraud, and improve accuracy. This study has used robust face detection algorithms, combined with Histograms of Oriented Gradients (HOG) feature extraction to yield a comprehensive database of authorized students face detection using Deep Learning (DL), when integrated with Principal Component Analysis (PCA), Support Vector Machine (SVM), KNearest Neighbor (KNN), and CNN classification, improves both system performance and accuracy. Enrollment generates unique identifiers, and regular updates to a centric dataset in order to make attendance tracking easier. This study promises to regulate attendance in schools in a practical and precise way.
- Open Access
- 1 July 2025
- Published by AMO Publisher in Scientia. Technology, Science and Society
- Vol. 2 (7) , 43-50
Automated Job Post Authenticity Verification
Due to recent advancements in social media and modern technologies, posting job openings online has become more prevalent. As a result, predicting deceptive job postings will be very important for everyone. Predicting fraudulent job listings encounters several obstacles similar to other classification tasks. This study suggests a machine learning method for identifying fraudulent job listings through a mix of text and category data. We gather different characteristics from the job listing, text, including the existence of specific keywords, along with elements from the job advertisement, such as the job position, type of employment, and necessary experience. Models such as Random Forest and Deep Neural Network are evaluated using different measures like accuracy, F1 score, ROC AUC score, and others following training. This study can be utilized to develop automated systems for identifying fraudulent job postings. Posts, aiding in safeguarding job seekers against scams and fraudulent actions in the employment market.
- 1 July 2025
- Published by AMO Publisher in Scientia. Technology, Science and Society
- Vol. 2 (7) , 43-50
Open AccessAutomated Job Post Authenticity Verification
Due to recent advancements in social media and modern technologies, posting job openings online has become more prevalent. As a result, predicting deceptive job postings will be very important for everyone. Predicting fraudulent job listings encounters several obstacles similar to other classification tasks. This study suggests a machine learning method for identifying fraudulent job listings through a mix of text and category data. We gather different characteristics from the job listing, text, including the existence of specific keywords, along with elements from the job advertisement, such as the job position, type of employment, and necessary experience. Models such as Random Forest and Deep Neural Network are evaluated using different measures like accuracy, F1 score, ROC AUC score, and others following training. This study can be utilized to develop automated systems for identifying fraudulent job postings. Posts, aiding in safeguarding job seekers against scams and fraudulent actions in the employment market.
- Open Access
- 15 April 2025
- Published by AMO Publisher in European Journal of Applied Science, Engineering and Technology
- Vol. 3 (3) , 27-34
Facial Feature Analysis for Habit Detection: Lip and Eye Colour in Smoking and Drinking Prediction
Analysing facial features now presents a real possibility for behaviour detection, such as habits prediction, especially on the prediction of activities like smoking and drinking. This review paper explores the utilization of lip and eye colour functionalities as important pointers in the detection of these behaviours through advancements made in image processing and machine learning. We evaluate the effectiveness of a range of computer methods for face cue extraction and interpretation in habitual behaviour prediction by surveying research from the last decade. In the review, issues such as dataset diversity, algorithmic correctness, and ethical constraints are covered while highlighting important correlations between smoking or drinking activities and differences in lip and eye colour. Our results open the door for further study in preventative care and individualized health monitoring by highlighting the potential of face feature analysis as a non-invasive diagnostic tool. This work synthesizes the body of current knowledge while pointing out potential and gaps for additional research in this multidisciplinary topic. In this research, we apply machine learning and image processing methods to explore whether analysis of facial features, specifically lip and eye colour, might be predictive of habitual behaviours such as drinking and smoking. We identify major developments, challenges, and possible directions for further work in this emerging area.
- 15 April 2025
- Published by AMO Publisher in European Journal of Applied Science, Engineering and Technology
- Vol. 3 (3) , 27-34
Open AccessFacial Feature Analysis for Habit Detection: Lip and Eye Colour in Smoking and Drinking Prediction
Analysing facial features now presents a real possibility for behaviour detection, such as habits prediction, especially on the prediction of activities like smoking and drinking. This review paper explores the utilization of lip and eye colour functionalities as important pointers in the detection of these behaviours through advancements made in image processing and machine learning. We evaluate the effectiveness of a range of computer methods for face cue extraction and interpretation in habitual behaviour prediction by surveying research from the last decade. In the review, issues such as dataset diversity, algorithmic correctness, and ethical constraints are covered while highlighting important correlations between smoking or drinking activities and differences in lip and eye colour. Our results open the door for further study in preventative care and individualized health monitoring by highlighting the potential of face feature analysis as a non-invasive diagnostic tool. This work synthesizes the body of current knowledge while pointing out potential and gaps for additional research in this multidisciplinary topic. In this research, we apply machine learning and image processing methods to explore whether analysis of facial features, specifically lip and eye colour, might be predictive of habitual behaviours such as drinking and smoking. We identify major developments, challenges, and possible directions for further work in this emerging area.
- Open Access
- 15 April 2025
- Published by AMO Publisher in European Journal of Applied Science, Engineering and Technology
- Vol. 3 (3) , 14-26
AI-Driven Hostel Monitoring System: Integrating Automated Attendance, Real-Time Visitor Detection, and Facial Recognition-Based Geo-Fencing Using
The existing hostel monitoring systems are based on flawed identification methods that can easily be spoofed, including GPS, Wi-Fi, RFID, or fingerprints. Our proposed approach addresses these constraints by utilizing face recognition-based geo-fencing, which will effectively and securely monitor movement inside hostel premises. Automatic attendance is implemented in real time, thereby avoiding human roll calls and maximizing efficiency. Moreover, the system can identify unauthentic people as well, hence preventing security violations through an integrated visitor detection module. Geo-fencing restricts entry and immediately informs the authorities in case of unlawful entry. The solution is highly accurate and scalable with the incorporation of hardware, such as cameras, edge computing devices, cloud storage, and real-time notifications powered by machine learning. Facial recognition and artificial intelligence make the hostels secure and efficient. In this paper, analysed the AI-based hostel monitoring systems, focusing on automatic attendance, real-time visitor detection, and face recognition-based geo-fencing. The research analyses the role of machine learning in enhancing security and efficiency in hostel management.
- 15 April 2025
- Published by AMO Publisher in European Journal of Applied Science, Engineering and Technology
- Vol. 3 (3) , 14-26
Open AccessAI-Driven Hostel Monitoring System: Integrating Automated Attendance, Real-Time Visitor Detection, and Facial Recognition-Based Geo-Fencing Using
The existing hostel monitoring systems are based on flawed identification methods that can easily be spoofed, including GPS, Wi-Fi, RFID, or fingerprints. Our proposed approach addresses these constraints by utilizing face recognition-based geo-fencing, which will effectively and securely monitor movement inside hostel premises. Automatic attendance is implemented in real time, thereby avoiding human roll calls and maximizing efficiency. Moreover, the system can identify unauthentic people as well, hence preventing security violations through an integrated visitor detection module. Geo-fencing restricts entry and immediately informs the authorities in case of unlawful entry. The solution is highly accurate and scalable with the incorporation of hardware, such as cameras, edge computing devices, cloud storage, and real-time notifications powered by machine learning. Facial recognition and artificial intelligence make the hostels secure and efficient. In this paper, analysed the AI-based hostel monitoring systems, focusing on automatic attendance, real-time visitor detection, and face recognition-based geo-fencing. The research analyses the role of machine learning in enhancing security and efficiency in hostel management.
- Open Access
- 1 November 2024
- Published by QTanalytics India (Publications)
- p. 95-106
Revolutionizing Examinations with the Ability Test Application
The Ability Test Application aims to transform online exams with an intuitive MCQ platform integrated into websites, offering automated scoring, comprehensive performance analytics, and secure test environments. Unlike conventional methods, which are labor-intensive and biased, this project supports hiring decisions with automated insights. In contrast to tools like ExamSoft and Google Forms, it enhances security and provides instant feedback. The project involves creating UI mock-ups, gathering requirements through interviews and surveys, and developing a reliable backend and responsive frontend. Rigorous testing ensures functionality, while advanced security features reduce human error and protect exam integrity. The application empowers hiring managers with data-driven insights and supports efficient decision-making by identifying trends and growth opportunities, ultimately redefining the examination process.
- 1 November 2024
- Published by QTanalytics India (Publications)
- p. 95-106
Open AccessRevolutionizing Examinations with the Ability Test Application
The Ability Test Application aims to transform online exams with an intuitive MCQ platform integrated into websites, offering automated scoring, comprehensive performance analytics, and secure test environments. Unlike conventional methods, which are labor-intensive and biased, this project supports hiring decisions with automated insights. In contrast to tools like ExamSoft and Google Forms, it enhances security and provides instant feedback. The project involves creating UI mock-ups, gathering requirements through interviews and surveys, and developing a reliable backend and responsive frontend. Rigorous testing ensures functionality, while advanced security features reduce human error and protect exam integrity. The application empowers hiring managers with data-driven insights and supports efficient decision-making by identifying trends and growth opportunities, ultimately redefining the examination process.
- Open Access
- 1 November 2024
- Published by QTanalytics India (Publications)
- p. 22-29
A LIME-based Explainable AI for Healthcare IoT: Building Trust in Clinical Decision-Making
The integration of Artificial Intelligence (AI) and Internet of Things (IoT) devices in healthcare offers vast potential for personalized medicine, remote monitoring, and early disease detection. However, complex Machine Learning (ML) models embedded in these systems often operate as "black boxes," hindering trust and transparency in critical medical decisions. Explainable AI (XAI) emerges as a key solution, aiming to demystify ML models and build trust in healthcare IoT applications. This paper explores the current challenges and opportunities in implementing XAI for healthcare IoT, proposing an architecture and methodologies for explainable clinical decision-making. We discuss promising XAI techniques, the integration of user interfaces for interactive explanations, and potential future directions for this crucial field.
- 1 November 2024
- Published by QTanalytics India (Publications)
- p. 22-29
Open AccessA LIME-based Explainable AI for Healthcare IoT: Building Trust in Clinical Decision-Making
The integration of Artificial Intelligence (AI) and Internet of Things (IoT) devices in healthcare offers vast potential for personalized medicine, remote monitoring, and early disease detection. However, complex Machine Learning (ML) models embedded in these systems often operate as "black boxes," hindering trust and transparency in critical medical decisions. Explainable AI (XAI) emerges as a key solution, aiming to demystify ML models and build trust in healthcare IoT applications. This paper explores the current challenges and opportunities in implementing XAI for healthcare IoT, proposing an architecture and methodologies for explainable clinical decision-making. We discuss promising XAI techniques, the integration of user interfaces for interactive explanations, and potential future directions for this crucial field.
- Open Access
- 1 January 2024
- Published by QTanalytics India (Publications)
- p. 113-122
Revolutionizing Elderly Care: Advanced Smart Fall Detection Solutions for Enhanced Safety and Independence
One of the most frequent reasons older individuals need medical attention is falling. Especially if they live alone, elderly individuals frequently hurt themselves from falls. In order to lower the danger of a victim, medical assistance must be given as soon as a fall happens. A number of systems have been created that use webcams to watch over the activities of senior citizens. But only indoor use is possible due to the high installation and running costs.The user of the currently available commercial product must wear a wireless wristwatchstyle emergency transmitter. Because of the device’s continuous swinging and moving, this strategy will limit user movement and increase the likelihood of false alarms.
- 1 January 2024
- Published by QTanalytics India (Publications)
- p. 113-122
Open AccessRevolutionizing Elderly Care: Advanced Smart Fall Detection Solutions for Enhanced Safety and Independence
One of the most frequent reasons older individuals need medical attention is falling. Especially if they live alone, elderly individuals frequently hurt themselves from falls. In order to lower the danger of a victim, medical assistance must be given as soon as a fall happens. A number of systems have been created that use webcams to watch over the activities of senior citizens. But only indoor use is possible due to the high installation and running costs.The user of the currently available commercial product must wear a wireless wristwatchstyle emergency transmitter. Because of the device’s continuous swinging and moving, this strategy will limit user movement and increase the likelihood of false alarms.