Automated Attendance Management System
0 likes21 views
This document presents a proposed automated attendance management system using face recognition. The system would use machine learning algorithms and deep learning approaches to recognize students' faces from images and track attendance. It discusses how face recognition works, including face detection, alignment, feature extraction, and recognition. It reviews similar existing systems and their limitations. The document tests several machine learning algorithms on their dataset and finds that an SVM classifier achieves the highest accuracy of 99.3%. Results are presented showing the system labeling and recognizing faces to mark attendance. The system aims to automate the attendance process to ease the burden on teachers, especially in online learning settings.
![International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 2516
Automated Attendance Management System
Shardul Khot1, Deepesh Shahdadpuri2, Mahek Tardeja3, Aakash Hotchandani4, Gaurav Tawde5
1-4Student, Dept. of Electronics and Telecommunication Engineering, Vivekanand Education Society’s Institute of
Technology, Maharashtra, India
5Assistant Professor, Dept. of Electronics and Telecommunication Engineering, Vivekanand Education Society’s
Institute of Technology, Maharashtra, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Almost everything that wedotoday isdoneusing
technology and is automated and linked online. The use of
internet of things, machine learning libraries and image
processing has been effectively used in almost all the fields. On
top of that the covid-19 pandemic hit us hard this year and
forced us to carry out learning in an online way, apart from
that some universities and colleges consider to continue with
far distance learning in the future even. We propose an
automated online attendance managementsystemthatwill be
based on face recognition. This is achieved using machine
learning algorithms and deep learning approaches after
initiating pre-processing of images of the students and then
training the model based on the face geometry. This
integration would make it easier for the teachers and staff to
manage classes in an automated way.
Key Words: Online attendance, Machine Learning, Face
Recognition, Deep learning, Eigen faces, SVM algorithm.
1. INTRODUCTION
Amidst the worldwide pandemic, a need to virtually carry
out all the traditional processes was mandatory, In Fact all
the schools, colleges and universities have adjusted to carry
out teaching, posting, collecting assignments and even
conducting exams through online platforms. In such a
situation it becomes very crucial to carry out all the
procedures in a traditional way as we used to do before the
pandemic, so with the help of our idea we have tried towave
off some of the burden and make things ultimately
simpler.
1.1 Implementation of Machine learning in facial
recognition technology:
As a rapid transformation has been witnessed in thefields of
AI, ML and deep learning technologies in the past few years,
this industry is rapidly progressing towards various
technologies and especially the growth factor in facial
recognition technologies has been tremendously good. To
get a proper gist of, what we can understand is that, this
technique is basically capable of recognizing a person based
on their different facial features. The technique of
recognition of facial features serves 4 vital purposes viz.
detection of all the faces, aligning themproperly,performing
feature extraction and finally recognizing them
Firstly, to locate the face in the image or video is
very crucial and until now most of the cameras
already have those built-in functions of detecting
the face accurately and thistechnologyissomething
that today even most of the social media platforms
also like Instagram, Snapchat, Facebook and many
more allow their users to add on various effects to
their pictures and videos.
Now coming to properalignmentoffacesthen,what
happens exactly is that faces are turned away from
the main point of focus so in such cases a ML
algorithm is trained in such a way that wherever
facial landmarks or features are marked out, the
desired results are obtained.
Moving further, then in this particular step what
will happen is it will cater into measuring and
extracting various different features and this
complete process is known as embedding, thereby
allowing it to distinguish the face from others.
Finally heading towards the last step, herein the
unique parameters of each face are measured and
with the help of a final machine learning algorithm
the measurements of the face are taken against
known faces present in a database.
2. LITERATURE REVIEW
In [1], researchers have made a complete web-based
application by producing better efficiency by integrating
various technologies. The authors have successfully
managed to build an existing system which is currentlyused
for managing attendancein Malaysia.Buttheefficiencycould
still be improved by using modern face recognition
algorithms to create a more robust system for achieving
better time complexity.
In [2], the authors have presented an automated attendance
system that uses biometrics. A fixed camera will capture
images and mark attendance which will be reflected in the
database. Also the system sends automated messages to
absent student’s parents. But this requires installation of
hardware.. For face recognition PCA algorithm is used but it
can be further improved by using LBP algorithm since it
makes the model more dynamic.](https://image.slidesharecdn.com/irjet-v9i4288-220927114919-0f6b6b6b/85/Automated-Attendance-Management-System-1-320.jpg)
![International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 2517
In [3], the author has proposed the method of face and head
detection in real time and has improved the time to
recognize and produce attendance records over 10 fps.This
involves the FDF technology which is a very famous and
robust method for real time face recognition as it involves
linear discriminant analysis by taking four directional
features of a face which could maintaina distinguishedarray
for each detected face. This method is very close to what we
are proposing as it does not involve contribution of any
hardware component. Therefore the algorithm works
uniquely firstly detecting skin color then morphology and
lastly labels it using fastly connected-component labeling
algorithm.
In [4] ,researchers have used a method of developing a
comprehensive embedded class attendance system using
facial recognition runs Raspbian (Linux) Operating System
installed on a micro SD card. Primarily the algorithm in the
proposed solution in Linear binary patterns. .The drawback
of the same system is that it involves a hardware system of
Raspberry Pi which itself cannot handle complex facial
recognition algorithms and therefore they have executed
with the local binary one which is not modern considering
the time complexity for greater samples of images.
In [5], the researchers of this paper have used Principal
Component Analysis abbreviated as PCA as an algorithm for
face recognition. The proposed solution is fast ,stable and
accurate and requires a computer with a camera only .
Principal component analysis is an older technique but still
considered efficient in the face recognition system for less
sample images with less computational speed.
3. MACHINE LEARNING ALGORITHMS:
A good accuracy score creates an impact that the ML model
is efficient enough to deliver the desired results. The
accuracy of any ML model depends either on its input or on
the data set used for training purposes.
Here we've used 3 different machine learning algorithms to
check our accuracy score which are as follows:
TRAINING MODEL ACCURACY SCORE
Logistic Regression 0.925
Random Forest 0.984
SVM classifier 0.993
4. CONCLUSION:
After implementing various machine algorithms and by
doing comparative analysis we can pitch that SVM classifier
algorithm gives us the highest accuracy score depending on
our dataset and it is also evident thatMachinelearning being
the future of coming world will nurture us with its amazing
and more efficient methods, solving the real worldproblems
on a much larger scale.
5. RESULTS:
The following are the results of our functioning system:
Fig 5.1 Face recognition and labeled image for 4 students.
Fig 5.2 CSV file showing attendance
Fig 5.3 Face recognition and labeled image for 13 students](https://image.slidesharecdn.com/irjet-v9i4288-220927114919-0f6b6b6b/85/Automated-Attendance-Management-System-2-320.jpg)
![International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 2518
Fig 5.4 CSV file showing attendance
REFERENCES
[1] M. Othman, S. N. Ismail and H. Noradzan, "An
adaptation of the web-based system architectureinthe
development of the online attendance system," 2012
IEEE Conference on Open Systems, 2012, pp. 1-6, doi:
10.1109/ICOS.2012.6417619.
[2] E. Varadharajan, R. Dharani, S. Jeevitha, B.
Kavinmathi and S. Hemalatha, "Automatic attendance
management system using facedetection," 2016Online
International Conference on Green Engineering and
Technologies (IC-GET), 2016, pp. 1-3, doi:
10.1109/GET.2016.7916753.
[3] Dr. A. Babu Karuppiah, M. Jeyalakshmi, L. Johnsilin
Shiny, B. Sri Devi, 2017, Online Attendance System,
INTERNATIONAL JOURNAL OF ENGINEERING
RESEARCH & TECHNOLOGY (IJERT) NCIECC – 2017
(Volume 5 – Issue 09).
[4] O. A. R. Salim, R. F. Olanrewaju and W. A. Balogun,
"Class Attendance Management System Using Face
Recognition," 2018 7th International Conference on
Computer and Communication Engineering (ICCCE),
2018, pp. 93-98, doi: 10.1109/ICCCE.2018.8539274.
[5] Chakraborty, Partha & Muzammel, Chowdhury &
Khatun, Mahmuda & Islam, Fahmida & Rahman, Saifur.
(2020). International Journal of Engineering and
Advanced Technology. 9. 93-99.
10.35940/ijeat.B4207.029320.](https://image.slidesharecdn.com/irjet-v9i4288-220927114919-0f6b6b6b/85/Automated-Attendance-Management-System-3-320.jpg)
Automated Attendance Management System
- 1. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN: 2395-0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 2516 Automated Attendance Management System Shardul Khot1, Deepesh Shahdadpuri2, Mahek Tardeja3, Aakash Hotchandani4, Gaurav Tawde5 1-4Student, Dept. of Electronics and Telecommunication Engineering, Vivekanand Education Society’s Institute of Technology, Maharashtra, India 5Assistant Professor, Dept. of Electronics and Telecommunication Engineering, Vivekanand Education Society’s Institute of Technology, Maharashtra, India ---------------------------------------------------------------------***--------------------------------------------------------------------- Abstract - Almost everything that wedotoday isdoneusing technology and is automated and linked online. The use of internet of things, machine learning libraries and image processing has been effectively used in almost all the fields. On top of that the covid-19 pandemic hit us hard this year and forced us to carry out learning in an online way, apart from that some universities and colleges consider to continue with far distance learning in the future even. We propose an automated online attendance managementsystemthatwill be based on face recognition. This is achieved using machine learning algorithms and deep learning approaches after initiating pre-processing of images of the students and then training the model based on the face geometry. This integration would make it easier for the teachers and staff to manage classes in an automated way. Key Words: Online attendance, Machine Learning, Face Recognition, Deep learning, Eigen faces, SVM algorithm. 1. INTRODUCTION Amidst the worldwide pandemic, a need to virtually carry out all the traditional processes was mandatory, In Fact all the schools, colleges and universities have adjusted to carry out teaching, posting, collecting assignments and even conducting exams through online platforms. In such a situation it becomes very crucial to carry out all the procedures in a traditional way as we used to do before the pandemic, so with the help of our idea we have tried towave off some of the burden and make things ultimately simpler. 1.1 Implementation of Machine learning in facial recognition technology: As a rapid transformation has been witnessed in thefields of AI, ML and deep learning technologies in the past few years, this industry is rapidly progressing towards various technologies and especially the growth factor in facial recognition technologies has been tremendously good. To get a proper gist of, what we can understand is that, this technique is basically capable of recognizing a person based on their different facial features. The technique of recognition of facial features serves 4 vital purposes viz. detection of all the faces, aligning themproperly,performing feature extraction and finally recognizing them Firstly, to locate the face in the image or video is very crucial and until now most of the cameras already have those built-in functions of detecting the face accurately and thistechnologyissomething that today even most of the social media platforms also like Instagram, Snapchat, Facebook and many more allow their users to add on various effects to their pictures and videos. Now coming to properalignmentoffacesthen,what happens exactly is that faces are turned away from the main point of focus so in such cases a ML algorithm is trained in such a way that wherever facial landmarks or features are marked out, the desired results are obtained. Moving further, then in this particular step what will happen is it will cater into measuring and extracting various different features and this complete process is known as embedding, thereby allowing it to distinguish the face from others. Finally heading towards the last step, herein the unique parameters of each face are measured and with the help of a final machine learning algorithm the measurements of the face are taken against known faces present in a database. 2. LITERATURE REVIEW In [1], researchers have made a complete web-based application by producing better efficiency by integrating various technologies. The authors have successfully managed to build an existing system which is currentlyused for managing attendancein Malaysia.Buttheefficiencycould still be improved by using modern face recognition algorithms to create a more robust system for achieving better time complexity. In [2], the authors have presented an automated attendance system that uses biometrics. A fixed camera will capture images and mark attendance which will be reflected in the database. Also the system sends automated messages to absent student’s parents. But this requires installation of hardware.. For face recognition PCA algorithm is used but it can be further improved by using LBP algorithm since it makes the model more dynamic.
- 2. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN: 2395-0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 2517 In [3], the author has proposed the method of face and head detection in real time and has improved the time to recognize and produce attendance records over 10 fps.This involves the FDF technology which is a very famous and robust method for real time face recognition as it involves linear discriminant analysis by taking four directional features of a face which could maintaina distinguishedarray for each detected face. This method is very close to what we are proposing as it does not involve contribution of any hardware component. Therefore the algorithm works uniquely firstly detecting skin color then morphology and lastly labels it using fastly connected-component labeling algorithm. In [4] ,researchers have used a method of developing a comprehensive embedded class attendance system using facial recognition runs Raspbian (Linux) Operating System installed on a micro SD card. Primarily the algorithm in the proposed solution in Linear binary patterns. .The drawback of the same system is that it involves a hardware system of Raspberry Pi which itself cannot handle complex facial recognition algorithms and therefore they have executed with the local binary one which is not modern considering the time complexity for greater samples of images. In [5], the researchers of this paper have used Principal Component Analysis abbreviated as PCA as an algorithm for face recognition. The proposed solution is fast ,stable and accurate and requires a computer with a camera only . Principal component analysis is an older technique but still considered efficient in the face recognition system for less sample images with less computational speed. 3. MACHINE LEARNING ALGORITHMS: A good accuracy score creates an impact that the ML model is efficient enough to deliver the desired results. The accuracy of any ML model depends either on its input or on the data set used for training purposes. Here we've used 3 different machine learning algorithms to check our accuracy score which are as follows: TRAINING MODEL ACCURACY SCORE Logistic Regression 0.925 Random Forest 0.984 SVM classifier 0.993 4. CONCLUSION: After implementing various machine algorithms and by doing comparative analysis we can pitch that SVM classifier algorithm gives us the highest accuracy score depending on our dataset and it is also evident thatMachinelearning being the future of coming world will nurture us with its amazing and more efficient methods, solving the real worldproblems on a much larger scale. 5. RESULTS: The following are the results of our functioning system: Fig 5.1 Face recognition and labeled image for 4 students. Fig 5.2 CSV file showing attendance Fig 5.3 Face recognition and labeled image for 13 students
- 3. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN: 2395-0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 2518 Fig 5.4 CSV file showing attendance REFERENCES [1] M. Othman, S. N. Ismail and H. Noradzan, "An adaptation of the web-based system architectureinthe development of the online attendance system," 2012 IEEE Conference on Open Systems, 2012, pp. 1-6, doi: 10.1109/ICOS.2012.6417619. [2] E. Varadharajan, R. Dharani, S. Jeevitha, B. Kavinmathi and S. Hemalatha, "Automatic attendance management system using facedetection," 2016Online International Conference on Green Engineering and Technologies (IC-GET), 2016, pp. 1-3, doi: 10.1109/GET.2016.7916753. [3] Dr. A. Babu Karuppiah, M. Jeyalakshmi, L. Johnsilin Shiny, B. Sri Devi, 2017, Online Attendance System, INTERNATIONAL JOURNAL OF ENGINEERING RESEARCH & TECHNOLOGY (IJERT) NCIECC – 2017 (Volume 5 – Issue 09). [4] O. A. R. Salim, R. F. Olanrewaju and W. A. Balogun, "Class Attendance Management System Using Face Recognition," 2018 7th International Conference on Computer and Communication Engineering (ICCCE), 2018, pp. 93-98, doi: 10.1109/ICCCE.2018.8539274. [5] Chakraborty, Partha & Muzammel, Chowdhury & Khatun, Mahmuda & Islam, Fahmida & Rahman, Saifur. (2020). International Journal of Engineering and Advanced Technology. 9. 93-99. 10.35940/ijeat.B4207.029320.