1.
Title – Smart Attendance System using Facial Recognition
2. Objective –
The objective of this project is to develop an automated attendance management system that uses facial
recognition technology to accurately identify individuals and record their attendance in real time. The
system aims to:
Eliminate manual errors and prevent proxy/buddy attendance.
Save time by automating the attendance process.
Ensure secure storage and easy retrieval of attendance records.
Provide analytics for monitoring attendance trends.
3. Input / Output of the System –
Input –
▪ Live Video Stream of Class
▪ Manual Upload Option – Images or Videos (in special cases, like rechecking
or unchecked attendance)
Output –
▪ Marked Attendance (Present/Absent/Latecomer)
▪ Attendance Reports – Daily, Weekly, Monthly
▪ Notifications – Alerts for Absentees / Low Attendance
▪ Teacher Dashboard – Attendance Records for Courses
4. Literature Review –
Face Recognition Smart Attendance System using Deep Transfer Learning – Khawla Alhanaee, Mitha
Alhammadi, Nahla Almenhali, and Maad Shatnawi (2021)
This study explores the use of transfer learning with CNN architectures such as AlexNet, GoogleNet,
and SqueezeNet for attendance systems. The results demonstrate high prediction accuracy with reduced
training times, showing CNNs as an effective approach for automating attendance compared to older
biometric methods like RFID and fingerprints.
Student Attendance with Face Recognition (LBPH or CNN): Systematic Literature Review – Andre
Budiman, Fabiana, Ricky Aryatama Yaputera, Said Achmad, and Aditya Kurniawan (2023)
This paper systematically compares LBPH and CNN algorithms for student attendance systems. Findings
indicate CNN consistently provides superior accuracy and robustness against external factors such as
lighting and pose variations, while LBPH remains useful in resource-limited scenarios due to its
simplicity and lower computational demand.
An Automated Face Detection and Recognition for Class Attendance – Chang-Horn Boe, Kok-Why Ng,
Su-Cheng Haw, Palanichamy Naveen, Elham Abdulwahab Anaam (2024)
This research employed HOG with SVM for face detection and CNN for recognition. The proposed
system was rigorously tested against QR-code-based attendance and was found to be more accurate and
efficient. It ensured reliable, real-time attendance data while minimizing manual errors and reducing
cheating opportunities
Smart Attendance System: Using Facial Recognition Technology – Amaan Ansari, Anita Pujari, and
Kamika Chaudhary (2024)
This paper presents a cloud-based solution integrating OpenCV, Python, and Firebase for real-time
attendance tracking. The system generates facial encodings, updates records automatically, and provides
monthly reports. It emphasizes scalability, high accuracy under varied conditions, and administrative
efficiency.
Automatic Attendance System Using Face Recognition with Deep Learning Algorithm – Ibrahim Al-
Amoudi, Rosdiyana Samad, Nor Rul Hasma Abdullah, Mahfuzah Mustafa, and Dwi Pebrianti (2022)
The authors design a cost-effective system based on MTCNN and FaceNet. Tested with over 1,000
images, the system achieved 100% detection and 87.03% recognition accuracy, proving the feasibility of
deep learning in real-world classroom environments without requiring additional hardware.
AttenFace: A Real-Time Attendance System Using Face Recognition – Ashwin Rao (2022)
This paper reports a system that takes classroom snapshots at fixed intervals instead of continuously
processing video streams. This reduced computational load while still ensuring reliable detection of
students’ presence throughout the class. The system also prevented proxy attendance and was designed
to integrate seamlessly with existing institutional platforms.
5. Target User –
This project’s primary target audience is educational institutions such as colleges and schools, where
manual attendance consumes significant time and is prone to errors or proxy marking. By automating
the process through facial recognition, the system ensures accuracy and efficiency in academic
environments.
While designed primarily for classrooms, the system also has room for modification to be adapted for
other domains such as corporate offices and governmental organizations, where attendance tracking and
workforce management are equally critical.
6. Impact to Society –
The Smart Attendance System using Facial Recognition has the potential to make a positive impact on
both academic and broader societal levels.
▪ For Educational Institutions: It reduces the time wasted in manual roll-calls, allowing faculty to
focus more on teaching and interaction. It also ensures fairness by eliminating proxy attendance,
which can affect overall student evaluation.
▪ For Students: The system builds accountability and discipline by encouraging punctuality and
reducing unfair practices.
▪ For Administration: Attendance records are automatically generated and securely stored,
improving transparency and reducing paperwork.
▪ Broader Societal Impact: With modifications, such systems can be extended to workplaces,
reducing administrative workload, enhancing workforce management, and promoting digital
transformation in governance and organizations.
▪ Sustainability Aspect: By digitizing attendance, institutions reduce dependency on paper
registers, contributing to environmentally friendly practices.
8. Uniqueness –
The proposed system is unique in that it is designed to automatically detect and recognize multiple faces
at once within a classroom environment, instead of requiring each student to manually scan their face
one at a time. This feature makes the system faster, more efficient, and better suited for real-world
classroom settings where time is limited and attendance must be recorded seamlessly.
Unlike conventional biometric systems such as fingerprint or RFID, which require individual interaction,
this project leverages facial recognition to capture attendance for an entire class simultaneously, thus
saving time and reducing disruption during lectures.
7. Timeline –
