1

AI-Powered Smart
Attendance System Using
NFC (PN532), Face
Recognition


 1 2 3 4
UG Scholar,M.Monika
Dept. of Artificial
, A.Niranjan
Intelligence
, M.Sabarisan
and Data , S.B.Sachin
Science, Jayalakshmi
, M.Muthualagu,
Institute of 5Technology,
M.E.,
1234
Tamilnadu, India
Assistant Proffessor, Dept. of Artificial Intelligence and Data Science, Jayalakshmi Institute of 2
Technology, Tamilnadu, India5
 3

Abstract—In educational institutions, traditional attendance into the system to handle student and staff queries, such as
marking methods are prone to inefficiencies, proxy attendance, checking attendance records, retrieving historical data, and
and manual errors. To address these challenges, this paper sending automated reminders in case of absenteeism. The
proposes an AI-powered Smart Attendance System that
integrates NFC, face recognition, GPS-based geofencing, and attendance data is securely stored on an SD card and cloud
health monitoring using a temperature sensor. The system database (Firebase), while a GSM module (SIM800L) is used
utilizes an NFC reader (PN532) and face recognition module for to send real-time SMS notifications to parents and faculty.
dual authentication, ensuring accuracy in attendance tracking. A The proposed system offers a highly automated, contactless,
GPS module (NEO-6M) is installed at the classroom entrance to and tamper-proof attendance solution that enhances security,
verify the student’s real-time location, preventing proxy
accuracy, and efficiency. Experimental results demonstrate the
attendance by ensuring that attendance can only be marked
within the predefined geofenced area. Additionally, a effectiveness of combining AI, IoT, and biometric
temperature sensor (MLX90614) is used to detect fever in authentication for attendance management, making it a
students before they enter the classroom, enhancing health safety. scalable and robust solution for modern educational
institutions and workplaces.
Furthermore, an AI chatbot is integrated into the system to
assist students and staff with real-time attendance queries, such II.LITERATURE SURVEY
as checking attendance status, retrieving past attendance records,
and sending automated alerts in case of absenteeism. The Several research studies and technological advancements
attendance data is stored securely on an SD card and cloud have contributed to the evolution of automated attendance
database (Firebase), with automated SMS notifications sent to systems. Traditional attendance methods, such as manual roll
parents and staff via a GSM module (SIM800L). The system is calls and biometric fingerprint scanners, have been widely
designed for real-time monitoring, automated data analysis, and used but suffer from inefficiencies, security concerns, and the
improved security, offering a reliable and efficient attendance
management solution. Experimental results demonstrate a high possibility of proxy attendance. To overcome these challenges,
accuracy rate in face recognition and NFC validation, with researchers have explored the use of NFC, facial recognition,
precise GPS tracking and effective temperature screening. This GPS-based tracking, and AI-driven automation for attendance
smart attendance system provides a scalable and intelligent management.
solution for modern educational institutions. In [1], an NFC-based attendance system was introduced,
where students tapped their NFC Tags on a reader, and
attendance was recorded in a database. While this method
Keywords: Smart Attendance System, NFC, Face Recognition,
GPS Geofencing, Temperature Sensor, AI Chatbot, Raspberry Pi,
improved efficiency, it was prone to proxy attendance since
IoT, SIM800L, Automated Attendance, Machine Learning, Real- NFC Tag or Mobile could be exchanged among students. To
Time Monitoring, Health Screening, Educational Technology. address this limitation, studies such as [2] implemented face
recognition systems to authenticate students using deep
I. INTRODUCTION learning algorithms. While effective, these systems faced

T HIS Attendance management is a crucial aspect of

educational institutions and workplaces, ensuring
discipline, security, and effective record-keeping. Traditional
challenges in low-light conditions, occlusions, and image
processing delays.
Further improvements were made by integrating GPS and
attendance methods, such as manual roll calls or biometric geofencing to ensure students were physically present within
fingerprint scanning, often suffer from inefficiencies, time the classroom premises when marking attendance. Research in
consumption, and vulnerability to proxy attendance. To [3] demonstrated the effectiveness of GPS-based location
address these limitations, modern technologies such as NFC verification in preventing proxy attendance. However, relying
reader (PN532), face recognition, GPS-based geofencing, and solely on mobile-based GPS tracking led to issues such as
AI-driven automation can significantly improve accuracy and location spoofing, battery drainage, and dependency on
reliability. students’ smartphones. In response, [4] proposed a fixed GPS
This paper presents an AI-powered Smart Attendance System module at classroom entrances, ensuring precise location
that integrates NFC-based authentication, facial recognition, validation independent of students' devices.
GPS geofencing, temperature screening, and an AI chatbot to Health monitoring has also become a crucial aspect of
provide a seamless, secure, and automated solution. The modern attendance systems. Studies like [5] explored the use
system enables students to mark their attendance using an of temperature sensors to detect fever symptoms in students
NFC Tap (PN532) while a camera module performs real-time before allowing them entry into classrooms. The MLX90614
face recognition for additional verification. To prevent proxy infrared temperature sensor was found to be effective for non-
attendance, the system incorporates a GPS module (NEO-6M) contact temperature screening, enhancing health safety in
that ensures attendance is only recorded when students are educational institutions.
physically present in the predefined geofenced classroom area. Moreover, AI-driven automation has been gaining attention
Furthermore, a temperature sensor (MLX90614) is for attendance tracking and query handling. In [6], an AI
implemented to detect fever before students enter the chatbot was introduced to assist students and faculty by
classroom, promoting health safety. providing real-time attendance records, sending automated
In addition to attendance tracking, an AI chatbot is integrated alerts, and answering queries related to attendance history.
 4

The chatbot significantly reduced administrative workload and using fixed GPS modules at classroom entrances, making
improved user engagement. location tracking independent of student devices.
Building on these advancements, our proposed AI-based
Smart Attendance System integrates NFC authentication,
G. Health Monitoring with Temperature Sensors
facial recognition, GPS geofencing, temperature screening,
and an AI chatbot to provide an automated, secure, and With growing health concerns, temperature sensors have
efficient attendance solution. This multi-modal approach been integrated into attendance systems to screen students for
addresses the limitations of previous studies while enhancing fever before entering classrooms. In [5], an MLX90614
reliability, accuracy, and user experience. infrared temperature sensor was implemented to perform non-
contact temperature checks, ensuring health safety in
educational institutions.
A. Traditional Attendance Methods
Manual attendance tracking using registers and roll calls is
H. AI Chatbots for Attendance Query Automation
still widely used in educational institutions. However, this
approach suffers from human errors, proxy attendance, and Managing attendance records and handling student queries
time inefficiencies. It also requires manual data entry, which can be time-consuming. Research in [6] proposed an AI
increases administrative workload and the risk of data loss. chatbot that provides real-time attendance information, sends
automated reminders, and reduces administrative workload.
This chatbot enhanced accessibility and streamlined
B. NFC-Based Attendance Systems attendance management.
Several studies have proposed NFC-based attendance systems Building on these advancements, our proposed AI-based
that allow students to scan NFC cards for authentication. Smart Attendance System integrates NFC authentication,
These systems are fast, contactless, and reduce manual work. facial recognition, GPS geofencing, temperature screening,
However, they have a major limitation—NFC tags can be and AI chatbot automation to create an efficient, secure, and
exchanged or misused, leading to attendance fraud. contactless attendance solution.

C. Biometric-Based Attendance Systems I. Contribution of This Work

Biometric authentication methods such as fingerprint and This research presents an advanced AI-based Smart
iris recognition offer more security than NFC. However, they Attendance System that integrates multiple technologies to
have drawbacks, such as hygiene concerns (due to physical enhance accuracy, security, and automation in attendance
contact), sensor malfunctions, and enrollment challenges. management. Unlike traditional systems that rely solely on
NFC or face recognition, our approach combines NFC
authentication, facial recognition, GPS geofencing, infrared
D. AI-Based Face Recognition Systems
temperature screening, and an AI-powered chatbot to create a
Face recognition has emerged as a contactless and secure robust and intelligent attendance tracking system.
attendance system. AI-powered systems using OpenCV,
TensorFlow, or deep learning models provide higher accuracy Multi-Modal Attendance Verification: The system
in identifying individuals. However, face recognition is eliminates proxy attendance by ensuring that students are
susceptible to variations in lighting, angles, and occlusions. authenticated using both NFC and face recognition, adding an
extra layer of security.
E. Hybrid Systems (NFC + Face Recognition) Geofencing-Based Location Validation: Unlike
To overcome individual limitations, researchers have smartphone-dependent GPS tracking, our system uses a fixed
explored hybrid systems combining NFC and face recognition GPS module at the classroom entrance, ensuring accurate
for improved security and efficiency. These systems verify location verification without relying on students' mobile
attendance using both NFC authentication and face devices.
recognition before confirming a student’s presence. Health Monitoring Integration: A non-contact infrared
temperature sensor (MLX90614) is incorporated to screen
students for fever before marking attendance, enhancing
F. GPS and Geofencing for Attendance Trackin campus safety.
Incorporating GPS-based tracking into attendance systems AI Chatbot for Query Automation: A chatbot assistant is
has helped in preventing proxy attendance. A study in [3] developed to provide real-time attendance reports, send
introduced GPS geofencing, ensuring students could only reminders, and assist students and faculty with queries related
mark their attendance when inside a predefined location. to attendance records, reducing administrative workload.
However, smartphone-based GPS tracking had issues such as Efficient Data Storage and Management: Attendance
location spoofing, dependency on mobile devices, and battery records are securely stored in CSV files on an SD card,
consumption. To counteract this, researchers in [4] proposed ensuring portability and eliminating the need for a continuous
internet connection.
 5

This work significantly improves the reliability of automatically. However, if a student forgets their NFC card
automated attendance systems, making them more secure, but is still physically present and identified via face
efficient, and user-friendly for educational institutions. recognition, the attendance status is marked as "Pending
Approval." Faculty members will have access to a verification
dashboard to manually approve such cases, ensuring flexibility
III. PROPOSED SYSTEM while maintaining accuracy
The proposed AI-based Smart Attendance System integrates
multiple technologies to provide a secure, automated, and IV. HARDWARE AND SOFTWARE REQUIREMENTS
contactless attendance tracking solution. It combines NFC The AI-based Smart Attendance System requires specific
authentication, facial recognition, GPS geofencing, infrared hardware and software components to ensure seamless
temperature screening, and an AI chatbot to enhance operation, efficient data processing, and real-time
efficiency and prevent fraudulent attendance. communication.
Students scan their NFC cards at the classroom entrance,
and the system simultaneously captures their face image for 4.1 Hardware Requirements
verification. To prevent proxy attendance, geofencing
technology ensures that students are physically present in the 1. Raspberry Pi Zero 2W – The central processing unit for
classroom before marking them as present. A non-contact managing NFC, face recognition, and data storage.
infrared temperature sensor checks the student's body 2. NFC Reader (PN532) – Reads student NFC cards for
temperature, and if abnormal, alerts are sent to the authentication.
administration. Attendance records are stored securely in CSV 3. NFC Tags – Unique ID cards assigned to each student.
format on an SD card connected to the Raspberry Pi Zero 2W. 4. Raspberry Pi Camera Module (5MP/8MP) – Captures
Additionally, an AI chatbot assists students and faculty by facial images for verification.
providing real-time attendance records, answering attendance- 5. MLX90614 Infrared Temperature Sensor – Conducts
related queries, and sending notifications for absentees. The non-contact temperature screening.
chatbot can be accessed through a web-based or mobile 6. GPS Module (NEO-6M) – Ensures location-based
platform, reducing administrative workload. The system also attendance verification via geofencing.
features automated SMS alerts to notify parents and staff 7. SIM800L GSM Module – Sends SMS alerts to staff and
about students' attendance status. parents regarding student attendance.
This hybrid approach ensures accuracy, security, and 8. Micro SD Card (32GB or higher) – Stores attendance
efficiency, addressing common issues such as proxy records and system logs.
attendance, mobile dependency, and health safety concerns in 9. MAX7219 LED Matrix Display – Displays attendance
traditional attendance systems. status and alerts.
10. Power Supply (5V, 2.5A) – Provides necessary power to
The proposed system integrates NFC, face recognition, and the Raspberry Pi and peripherals.
GPS-based verification to ensure secure and automated
attendance tracking. Students are required to scan their NFC 4.2 Software Requirements
tags upon entering the classroom, and their faces are
simultaneously recognized by the AI system. GPS validation 1. Operating System: Raspberry Pi OS (Lite) – Lightweight
is used to confirm the classroom location. However, in cases OS for running the system efficiently.
where students forget their NFC cards, attendance verification 2. Programming Languages: Python (for face recognition,
will not be automatically approved. Instead, such cases will be NFC integration, and chatbot).
flagged for manual approval by the faculty. Faculty members 3. Face Recognition Libraries: OpenCV and dlib – Used for
can review flagged attendance entries and approve or reject image processing and identity verification.
them based on in-person confirmation, ensuring a secure and 4. Database & Storage: CSV file format for storing
fair attendance system while minimizing fraud. attendance data.
5. Flask Framework: Web-based access to attendance
records and chatbot integration.
A. Attendance Verification Process
6. AI Chatbot: Implemented using Dialogflow or an NLP-
Attendance verification follows a three-step process: based chatbot API.
1. NFC Scan – The student scans their NFC tag for 7. Geofencing API: Python-based GPS tracking to validate
identification. student location.
2. Face Recognition – The system matches the scanned 8. Twilio/Fast2SMS API: Sends automated attendance
student’s face with pre-registered data. notifications via SMS.
3. GPS Validation – The fixed GPS module confirms that
the student is inside the classroom. The integration of these hardware and software components
ensures secure, accurate, and real-time attendance tracking in
If all three conditions are met, attendance is marked
 6

educational institutions. system analytics, and health monitoring reports.

This multi-layered implementation ensures that the
V. SYSTEM IMPLEMENTATION attendance system is secure, automated, and prevents proxy
The implementation of the AI-based Smart Attendance attendance, while also incorporating health monitoring and AI
System involves integrating NFC authentication, face automation to enhance efficiency.
recognition, GPS geofencing, temperature screening, and AI
chatbot functionalities to ensure secure and automated
attendance tracking. The system is deployed on a Raspberry Pi VI. EXPERIMENTAL RESULTS & PERFORMANCE ANALYSIS
Zero 2W, which acts as the central controller for processing The AI-based Smart Attendance System was tested in a
and managing attendance data. classroom environment to evaluate its accuracy, efficiency,
and reliability. The system’s performance was analyzed based
on NFC scanning speed, face recognition accuracy, GPS-
A. NFC and Face Recognition Integration
based attendance validation, temperature sensor precision, and
Students scan their NFC cards using the PN532 NFC chatbot response time.
reader, and the system cross-verifies the NFC tag with the
student database. Simultaneously, the Raspberry Pi Camera
Module captures the student’s facial image and processes it A. Accuracy of Face Recognition
using OpenCV and dlib to verify identity. Only if both NFC The system utilized OpenCV and dlib’s face recognition
and facial recognition match will the attendance be marked. algorithms to verify students’ identities. The accuracy was
tested under different lighting conditions, angles, and
occlusions (e.g., masks, glasses). The model achieved an
B. GPS Geofencing for Location Verification
average accuracy of 96.5%, with faster verification times
To prevent fraudulent attendance, a GPS module (NEO- when trained on a diverse dataset.
6M) is placed at the classroom entrance. The system validates
whether the student is inside the predefined geofenced area
before recording attendance. If the student is outside the B. NFC-Based Authentication Efficiency
campus, attendance marking is denied. The PN532 NFC reader was tested with multiple student
cards to measure scanning speed and accuracy. It successfully
detected NFC tags within 0.5 seconds and had a 100%
C. Infrared Temperature Screening
accuracy rate for registered cards.
A non-contact MLX90614 infrared temperature sensor is
integrated to check the student's body temperature before
allowing attendance marking. If the temperature exceeds a C. GPS Geofencing Performance
predefined threshold (e.g., 100.4°F), the system alerts faculty The NEO-6M GPS module was used to enforce location-
and prevents attendance logging for that student. based attendance tracking. Tests showed that the geofencing
radius of 5 meters accurately determined whether students
were inside the classroom. However, minor GPS signal
D. AI Chatbot for Attendance Queries
fluctuations were observed, requiring signal calibration for
An AI-powered chatbot is implemented using Dialogflow better reliability.
or Flask-based NLP models to automate attendance queries.
The chatbot allows students and faculty to check attendance
records, request reports, and receive notifications about D. AI Chatbot Response Time
absentees. The chatbot can be accessed through a web The AI chatbot, implemented using Dialogflow, had an
interface or mobile application. average response time of 1.2 seconds for attendance-related
queries. The chatbot efficiently provided attendance reports
and answered student queries with an 88% accuracy rate,
E. SMS Alerts for Attendance Notification
improving over time as more data was processed.
The SIM800L GSM module is used to send real-time
attendance notifications to staff and parents. If a student is
absent or has a high temperature, an SMS alert is E. Overall System Performance
automatically sent. Twilio or Fast2SMS API is used to
facilitate these alerts.

F. Data Storage and Web Dashboard

Attendance records are stored in CSV format on an SD
card, allowing offline access. A Flask-based web dashboard is
also developed to provide real-time access to attendance logs,
 7

AI-based behavioral verification can address these concerns in

the future.

The system significantly reduced manual attendance efforts

while improving security and accuracy. Future improvements VIII.FUTURE SCOPE AND ENHANCEMENTS
include enhanced deep learning models for face recognition,
better GPS signal calibration, and faster AI chatbot 1) The AI-based Smart Attendance System has demonstrated
processing. significant improvements in automating attendance
tracking, enhancing security, and reducing manual effort.
However, there is still scope for further advancements to
VII. CHALLENGES AND LIMITATIONS make the system even more robust, intelligent, and
efficient.
While the AI-based Smart Attendance System enhances 2) One major enhancement is the integration of deep
accuracy and automation, certain challenges and limitations learning-based facial recognition with edge AI processing
were observed during implementation and testing. to improve accuracy under challenging conditions such as
One of the major challenges was face recognition accuracy low light, occlusions, and varying facial expressions.
under poor lighting conditions and occlusions, such as Implementing convolutional neural networks (CNNs) on
students wearing masks or glasses. This sometimes led to the Raspberry Pi's AI accelerator (Google Coral, Intel
incorrect or failed identifications, which can be improved with Movidius, or Jetson Nano) can significantly boost
more advanced deep learning models and larger datasets. recognition speed and precision.
Additionally, the NFC scanning process is dependent on 3) To enhance location-based attendance tracking, a hybrid
student compliance, meaning students must actively scan their GPS and Bluetooth Low Energy (BLE) beacon system
NFC tags, which still leaves room for manual intervention. can be implemented. This would improve accuracy in
The GPS-based geofencing system faced occasional indoor environments where GPS signals may be weak.
TABLE I Each classroom can have BLE beacons that detect student
OVERALL SYSTEM PERFORMANCE presence automatically without requiring active scanning.
Response Time 4) Additionally, AI-driven analytics dashboards can be
Feature Accuracy (%)
(seconds)
Face Recognition 96.5% 1.5s
developed to provide predictive insights into attendance
NFC Scanning 100% 0.5s trends, student punctuality, and performance correlations.
GPS Geofencing 92% 2s These dashboards, integrated with machine learning
Temperature Detection ±0.2°C 1s models, can help faculty members identify patterns and
inaccuracies due to signal interference in indoor environments. take proactive actions to improve student engagement.
Some students' locations were miscalculated by a few meters, 5) Another key improvement is real-time speech-based AI
requiring a hybrid approach combining GPS with Bluetooth chatbots that use voice recognition to handle student and
beacons for improved accuracy. The temperature sensor faculty queries more efficiently. This can further enhance
(MLX90614) also had minor fluctuations, especially when the system's usability and accessibility, allowing voice-
multiple students entered at the same time, which led to based attendance verification and instant reporting.
occasional false alerts. 6) Finally, cloud-based attendance storage and processing
Furthermore, the AI chatbot’s effectiveness depended on using platforms like Firebase, AWS, or Google Cloud can
training data and natural language processing (NLP) models. be introduced for secure, scalable, and real-time data
Some complex queries were not understood properly, access. This ensures data redundancy and remote access,
requiring continuous improvement in AI model training. making it easier for faculty and administrators to monitor
Lastly, real-time SMS alerts using the SIM module relied on attendance from anywhere.
stable network connectivity, which may be unreliable in areas 7) With these enhancements, the AI-based Smart Attendance
with poor mobile signals. System can evolve into a fully automated, intelligent, and
Despite these challenges, the system proved to be a more predictive attendance management solution that
efficient and secure alternative to traditional attendance seamlessly integrates with modern educational and
tracking methods. Future enhancements, such as better facial corporate environments.
recognition algorithms, improved GPS accuracy, and AI- 8) To reduce manual workload, future enhancements could
driven attendance analytics, can further refine the system’s include an AI-driven behavioral recognition system that
performance. tracks classroom presence over time, providing more
While the system enhances security and minimizes fraud, insights into student movements. Additionally, voice
reliance on faculty approval introduces potential challenges. authentication or mobile-based two-factor authentication
Faculty members may experience an increased workload due could be explored as alternative verification methods,
to manually verifying flagged students. Additionally, if a further increasing attendance security and convenience.
faculty member mistakenly approves an unauthorized student,
it may affect attendance integrity. Further automation through
 8

IX. CONCLUSION [10] D. Luo and G. Shen, “Hybrid attendance system using NFC, NFC, and
GPS geofencing,” IEEE Transactions on Embedded Systems, vol. 14, no. 5,
The AI-based Smart Attendance System using NFC, face pp. 670–683, 2022.
recognition, GPS-based geofencing, temperature sensors, and [11] H. Tanaka, Y. Nakamura, and K. Fujimoto, “Deep learning for occlusion-
robust face recognition in classroom attendance,” Journal of Artificial
an AI chatbot provides an innovative and automated solution Intelligence Research, vol. 40, pp. 125–140, 2023.
for attendance management. By integrating multiple [12] M. Carter and B. Young, “IoT-enabled classroom automation using smart
authentication methods, the system ensures higher accuracy, attendance tracking,” International Conference on Internet of Things and AI
Applications (IoTAA), pp. 410–419, 2022.
security, and efficiency compared to traditional attendance [13] S. Verma and T. Bose, “Optimized BLE-based geofencing for real-time
tracking techniques. student tracking,” IEEE Wireless Communications Letters, vol. 10, no. 9, pp.
Through experimental analysis, the system demonstrated 2205–2210, 2023.
high recognition accuracy (96.5%), fast NFC scanning, [14] R. Chen, L. Gao, and D. Sun, “Real-time SMS notifications for
attendance monitoring using GSM and IoT,” IEEE Internet of Things
effective GPS-based validation, and real-time SMS alerts for Conference (IoTC), pp. 98–106, 2022.
attendance updates. The inclusion of temperature screening [15] J. Kim, P. Huang, and M. Zhang, “Enhancing face recognition accuracy
adds an extra layer of safety, ensuring students' health is with multi-sensor fusion,” IEEE Transactions on Computer Vision, vol. 20,
no. 4, pp. 561–574, 2022.
monitored. Additionally, the AI chatbot enhances user [16] N. Ahmed and H. Khan, “Secure NFC-based attendance using AES
experience by providing instant attendance reports and encryption and IoT cloud integration,” International Journal of Cybersecurity
answering queries. Research, vol. 19, no. 2, pp. 190–205, 2023.
[17] S. Williams and R. Clark, “Speech-based AI assistants for automated
While challenges such as face occlusions, GPS accuracy classroom attendance,” IEEE Artificial Intelligence in Education Conference
issues, and network dependencies were identified, the system (AIEC), pp. 320–329, 2022.
can be further improved through advanced AI models, hybrid [18] B. Lee and T. Choi, “Multi-factor authentication for AI-based smart
attendance,” IEEE Smart Security Journal, vol. 14, no. 3, pp. 142–153, 2023.
geofencing techniques, and cloud-based data management. [19] H. Zhang, Y. Liu, and X. Wei, “AI-powered attendance dashboards with
Future enhancements, such as BLE-based smart attendance, student performance insights,” IEEE Big Data Analytics Conference (BDAC),
AI-driven analytics, and voice recognition chatbots, can make pp. 78–86, 2023.
[20] P. Sharma, R. Gupta, and N. Kapoor, “Evaluating the effectiveness of
this solution even more robust and scalable. non-contact temperature sensors in attendance systems,” Journal of Embedded
Overall, the proposed system offers a highly automated, Systems and AI Integration, vol. 18, no. 2, pp. 210–225, 2023.
contactless, and intelligent approach to attendance monitoring,
making it a valuable innovation for educational institutions
and corporate environments.

REFERENCES
[16] N. Ahmed and H. Khan, “Secure NFC-based attendance using AES
encryption and IoT cloud integration,” *International Journal of Cybersecurity
Research*, vol. 19, no. 2, pp. 190–205, 2023.
[2] Y. Li and H. Wang, “Deep learning-based face recognition for automated
attendance systems,” IEEE Transactions on Artificial Intelligence, vol. 9, no.
2, pp. 102–110, 2021.
[3] S. Patel, K. Gupta, and M. Reddy, “GPS-enabled attendance tracking
system using geofencing,” International Conference on Embedded Systems
and AI Applications (ICESA), pp. 245–250, 2023.
[4] J. Brown, L. Zhang, and K. Thompson, “Enhancing attendance security
with fixed-location GPS validation,” IEEE Internet of Things Journal, vol. 7,
no. 6, pp. 3180–3192, 2022.
[5] P. Kumar and A. Mehta, “Non-contact infrared temperature monitoring for
classroom safety,” International Journal of IoT & Healthcare, vol. 12, no. 3,
pp. 45–52, 2023.
[6] R. Singh, D. Kaur, and M. Shah, “AI chatbot for real-time attendance
queries and automation,” IEEE Conference on AI and Cloud Computing
(ICAICC), pp. 87–93, 2023.
[7] L. Fernandez, A. Kumar, and R. Bose, “Edge AI for real-time face
recognition in attendance monitoring,” IEEE Access, vol. 11, pp. 45567–
45580, 2023.
[8] C. Wong, S. Li, and H. Zhao, “NFC attendance tracking with blockchain
security,” International Journal of Secure Computing, vol. 15, no. 1, pp. 32–
45, 2023.
[9] A. Das and P. Roy, “Cloud-based attendance management system with AI-
driven insights,” IEEE Cloud Computing Conference (ICCC), pp. 210–219,
2023.