AI-Based Attendance Monitoring System
AI-Based Attendance Monitoring System
AI-Based Attendance Monitoring System
net/publication/338331249
Conference Paper in International Journal of Innovative Technology and Exploring Engineering · January 2020
DOI: 10.35940/ijitee.B1057.1292S19
CITATIONS READS
3 16,133
4 authors, including:
Rashmi Krishnappa
GITAM University
2 PUBLICATIONS 4 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Naveen Reddy K P on 01 January 2020.
Abstract: Attendance Monitoring System is essential in all In recent years we observed remarkable changes in face
organizations for checking the performance of students and it is recognition techniques because of available biometric
not easy task to check each and every student is present or not. In methods, this is the most unnoticeable technique. The
all organization attendance are taken manually by calling their installation face recognization systems on a large scale are
register numbers or names and noted in attendance registers
easy but the actual implementation of face recoginization
issued by the department heads as a proof and in some
organizations the students wants to sign in these sheets which system is ambitious because it has to take into account for
are stored for future references. This technique is repetitive, all potential cases variation caused by a modification in face
complex work and leads to errors as few students regularly sign expressions by light-weight, face expressions, different
for their absent students or telling proxy attendance of the absent styles, image resolution, sensing element device, viewing
students. This method additionally makes it more complex to distance, etc.[3] Several algorithmic rules are implemented
track all the students attendance and difficult to monitoring the on face recognition and every algorithm has strengths and
individual student attendance in a big classroom atmosphere. In capabilities by its own. We tend to do face recognition
this article, we use are using the technique of utilization face nearly on a daily basis. Most of the time we glance at a face
detection and recognition framework to contunisuly recognize
and acknowledge by in a flash with the data present already
students going to class or not and marking their attendance by
comparing their faces with database to match and marking in the database
attendance. This facial biometric framework takes a picture of a This aptitude if potential followed by machines will
person using camera and contrast that image and compare the influence be valuable and should give for a vital role in real-
image with the image with is stored at the time of enrolment and world applications like vary access management, national
if it matches marks the attendance and monitor the student and international security and defense, etc. At present
performance contunisuly. We may use the concept of artificial mainly two approaches rely on Face reorganization
intelligence concept to monitor student attendance like capturing methods. The first and very familiar method is native face
the motion pictures of the student when present in class to recognition system that depends on face expressions of a
analyze the student data how much time the student presents in
covering for example eyes, nose, color, etc. to identify the
class.
face with a someone matching or not. The second approach
Index Terms: Artificial Intelligence, Student Attendance is world face recognition system which uses the whole face
System, Face reorganization, Students attendance monitoring to identify a person. The two described approaches are
system and applications enforced by a methods by various types of algorithms. The
recent implementation using artificial intelligence
I. INTRODUCTION: applications in face recognition attracts many scientists to
research on this topic. The elaboration of a face features
The student's attendance system using artificial originates from the changes continuously within the facial
intelligence concept mainly works using the concept of expression that changes by time. Apart all these changes, we
facial recognization system was discussed by Akshara are ready to acknowledge an individual easily.The idea of
Jadhav et al.0.Face is considered as a primary key feature to developing a self understanding and self-learning intelligent
identify and talk with other peoples in the world because machine may require to give sufficient data to the machine
face considered as a unique identity for each and every was proposed by Pradeepa .M et al.[5] Facial Recognition
person. The facial features will be unique to the other System can be defined in simple words as the technology
indusial. The Unique features for every indusial make facial that identifies a person and verifies it with the database by
recognition in implementing the real world. comparing the facial features described by Chaitanya reddy
Human distinguish a particular persons face based on [7].
several factors like color, nose, eyes, ears, etc but for Face Recognition :
computers, it's difficult to analyze the data so we may use
the concept of Computer vision. The intention of using Scientists started working on computers to recognize human
computer vision technology to recognize the human features faces from mid-1900s because of its enormous applications
in a computer. on face recognition has received continuous attention from
researchers. Face recognition may be outlined because of the
technique of characteristic by someone based on biometrics
by the approach of matching a capturing image or video
with the data present in the database. The data flow process
in face Recognition systems starts by having the ability to
Revised Manuscript Received on December 15, 2019. find face and recognize frontal faces from data input devices
K. P. Naveen Reddy, GITAM School of technology, Bangalore like mobile phones, cameras, etc. [10] Practically it has
campus. been proven that students attended classes only when there
T. Alekhya, GITAM School of technology, Bangalore campus. is full control on classroom and attendance monitoring.
Sushma Manjula T, GITAM School of technology, Bangalore
campus.
Ms. Rashmi K, Assistant Professor, Department of CS&E, GITAM
School of Technology, Bangalore
Published By:
Retrieval Number: B10571292S19/2019©BEIESP Blue Eyes Intelligence Engineering
DOI: 10.35940/ijitee.B1057.1292S19 592 & Sciences Publication
AI-Based Attendance Monitoring System
II. METHODOLOGY
A. Functional Specifications:
Functional specifications are the requirements in which
requires to operate a system. These requirements are
necessary to assemble a system which will be required to
attain the objectives embarked on previously was
implemented by Jireh Robert Jam[3] .Some of the important
functional and non functional requirements are outlined
below by analyzing the shop keeper story.
o First capturing the facial image by high qualitity camera
o HD Camera especially professional camera
o Facial features should be detected in photo.
o Crop the overall ranges of faces detected. Fig-1: Flow Diagram describing the face detection.
o Resize all the images until the recognization system takes From the above use-case mentioned Face Detector detects
photo to recognize. the face inputted in the given image or video and loads the
o Calculating the overall attendance percentage based on image to the system. Face localization can notice wherever
facial features matched. face is situated within the inputted image or video marked
o Storing all the detected face images in a folder. using the bounding boxes. The landmarks of the face like
o Loading the images into the database. eyes, color, nose, mouth etc are done for feature extraction
o We want to train facial features to computer to recognize. from the system using face localization. The key features are
o Perform recognition for faces stored on database. extracted using face extraction to undergo tracking
o Calculate computer facial recognization speed for effecetively.Facial features are matched and classify with
effective security. the databases stored. Face detection offers two outputs either
o Performing face recognition sequentially for the each a positive output or negative output for the supported image
image cropped. from the set of stored database collection of images.
o Displaying input and output cropped image side by side
on a same slot to recognize and compare the features byC. Working of face reconization system
machine. The working of face recoginization system mainly classified
o After recognizing the face displays the name of the output into two types of algorithms [3]. They are holistic matching
image above the image in the given area to identify easily. algorithm and feature-based algorithm The entire face is
considered as input data to identify a particular person
B.Non-Functional Specifications:
from the database in the holistic matching method and the
Non-functional Specifications are the needs based on the face is divided based on the facial features like eyes, color,
specific criteria to evaluate the operation of the system. skin tone and eyebrows etc in feature-based method. Apart
These requirements are collect and analyzed based on the from the above two, in recent days three dimensions face
client needs and exceptions, security and working etc recoginization technique is used to capture the 3D faces
implementation issues by Jireh Robert Jam[3] Specifically: using sensors to identify person in real world very
o The first and important thing is user need to find easy to accurately. Real-time face detection using 3D sensor based
take pictures. applications determine a person with various facial images
o The system can be easily installed. in different angles, on different light conditions, and various
o The operator will give a clear instruction on how to pose poses with different expression makes more efficient in all
the face to train computer. applications.
o The face reconization system is highly secure.
o The response time if the system is very less i.e. 10 III. ARTIFICIAL INTELLIGENCE TO MARK
seconds. ATTENDENCE:
o The face recoginization system must be fast, reliable and Artificial Intelligence enabled Face detection based
100% efficient. application becoming world famous. This was discussed by
Santana Fell [6]. In Washington, American private
elementary school in Seattle begins implementing automatic
face detection technique called SAFR(Secure accurate
Facial Recognition). In this method faces of parents are
adding into the database.
Published By:
Retrieval Number: B10571292S19/2019©BEIESP Blue Eyes Intelligence Engineering
DOI: 10.35940/ijitee.B1057.1292S19 593 & Sciences Publication
International Journal of Innovative Technology and Exploring Engineering (IJITEE)
ISSN: 2278-307A5, Volume-9 Issue-2S, December 2019
The security gates of the school opened only Secure The personal information in mobiles can be protected in
accurate Facial Recognition recognizes the right person. The smart phones to check no one can access the personal
working of this method is based on scanning These method data even the smart phone is stolen. It is being used in
works by the method of scanning the persons face and check mobile phones for unlocking. This method is very high
with the data base and the person need not to show their face secure.
to scanner it will automatically detects face and check with
A. Advertising makes more responsive
database if both matches the doors will open
Advanced Artificial intelligence enabled techniques are Using face identification method we can makes
started implemented in class rooms. As books are replaced advertisements a lot of participating and makes more
by tablets and smart board's blackboards..Smart mobiles will personalized for various types of users. To customize the
replaces attendance registers place. A girls government high audience interests Some branded companies has already
school in Chennai, Tamil Nadu, started implementing implemented automatic face identification method in digital
attendance system based on artificial intelligence enabled world to customize their campaigns by scanning a face,
and this is a India first school implemented and got popular based on his age and gender ads are played. Apart that this
in Japan and Us .The working of that method is based on the system identifies expressions of people to understand their
mobile application in which contains all the relevant basic emotions like sad, happy, disgusted, etc. based on that
details of the student with the student photo. The faculty displays adds for a particular product in which the user likes
needs to just click the photo which automatically marks the by understanding the facial emotional features.
attendance those who present and stores the attendance in B. Airport security increases
database. Airports are the most busy place in which high chances of
A Company in china has programmed the world's 1st criminal and terrorists activates because of this reason
Artificial Intelligence teaching assistant designed to control several airlines started implementing face identification
through whiteboard along with a commanding person. systems to check baggage’s and flight boarding makes the
Artificial Intelligence assistant, reacting to the user through process quicker. Moreover, the Artificial Intelligence face
voice recognition. The main feature of automatic face authentication application implemented with surveillance
recognition is to find that all students attending or not. It is cameras which helps in identifying a terrorist who might be
also used to mark grades and use to prepare customizes involved in some disastrous activity by understand the
exercises for the college activities. unusual behavior the person and the facial expressions of the
Citytech Software firm in Calcutta organized an in-house person to recognize the criminals to make the airport place
innovation competition based on trending technology. One safe.
group in that contest started working on Artificial
Intelligence enabled cameras to require pictures of C. Diagnose rare genetic diseases
individuals coming into the workplace notice the person age, Artificial Intelligence enabled automatic face identification
feelings ,gender and alerts the concern person about the application will facilitate the medical business to diagnose
security thread discussed by Shreyak Sawhney et al.[9] The sickness that leads to an amendment in appearances like
CCTV cameras which detects the employee face and marks spreading eyes or drooping ears. A face recognition
the attendance of that person and calculates their salary scanning will becoming a part of standard medical check-
based on how many hours they worked in office. up that will identify genetic disorders such as Disgorge
syndrome, Angelman syndrome, Cornelia de Lange
IV. REAL-WORLD APPLICATIONS OF FACIAL syndrome etc which bring gradual changes in facial
RECOGNITION: expressions. Now clinical diagnoses for various genetic
Recent Years, Face Recognition and identfication is diseases and its treatment will become faster than before
being extensively used in security surveillance Real- with face recognition in this way facial technology also
time system to Identifying individuals on the spot implemented in medical field also.
proposed by Shreyak Iyer et al [8].
D. Provide driver safety and personalization
In crime detection and forensic analysis it’s plays a
major role. Using drivers license to identify the Automobile companies like Tesla, Subaru etc are increasing
criminals using facial recognition system is used in US their services in numerous ways in which by utilizing face
Federal Bureau of Investigation. Artificial Intelligence recognition systems to recognize drivers. The main use of
enabled cameras has been checked to identify those using face recognization system is to begin the car using
smuggling persons in UK. face recognization rather than using key to start the car. Face
Face identification method plays a key role in making recognition can scan the facial features of the driver and
secure payments using online payment. online monitors the focus of the driver and alerts the concern
payments as it is more trusted feature in which only the person if they are losing the
account holder can access account. concentration.
Published By:
Retrieval Number: B10571292S19/2019©BEIESP Blue Eyes Intelligence Engineering
DOI: 10.35940/ijitee.B1057.1292S19 594 & Sciences Publication
AI-Based Attendance Monitoring System
The face identification identifies the preference of the driver effect and reducing the cost of employees too. Then any
like favorite stations, calendar and the position of the seat organization can recognizes the fact that usage of automated
with detailed report which increases the customer and driver face identification is highly secure with accurate data.
safety.
D. Forget the Time Fraud
E. Helpful in VIP identification The massive advantage using automatic face recognition
A face recognition system will establish prestigious guests method is to provide the time tracking attendance system to
whereas getting into a building or attending any event enable avoid time fraudness between employees. It is not
automatically identifies the person which boosts their possible to any colleagues to favor their friends because
loyalty greatly. In hospitals, the face recognition method everyone need to pass the entrance gate where the face
helps in identifying a returning guest at the entrance and recoginization camera catches their photo and matches with
displays the preference of food; room etc once the face is the database and avoids time fraud among workers. This is
scanned. In event the organizer can easily identifies the VIP very beneficial for the employees who works based on
guest among all the fans early and skips the queue and hourly starts check in the time starts counting from that
provide all VIP benefits to them.. moment until a similar check-out done which is benefit to
the organization they need not to monitor their workers and
F. Stops retail crime
the method is quick because of the actual fact that staff
In retail business, a face recognition and identification doesn’t have to prove indusial identities by scanning their
system could be a future decider because it identifies the smart cards on the scanner. It is very difficult for the
person instantly once the person enters it searches for a business heads to monitor all the employees are attending or
thief, criminal or person with fraudulent history. The not. The main problem is time fraud among the employees
security officer of the particular retail shop informed can be avoided using artificial intelligence enabled face
immediately when the criminal enters. When a criminal detection system.
enters into store and the image of criminal is matching with
the huge database of criminals for any criminal records VI. LIMITATIONS:
pending. With the help of face recognition technology the
A. Processing & Storing
retail crimes are gradually decreasing and there is no case of
retail crimes and shop robberies etc. Storages plays major role in practical world in avoiding
wasting huge amounts of information and needs to store for
V. ADVANTAGES: future use. For storing HD-Videos in a very low resolution
also needs a huge amount of .We are wasting large number
A. Improvement of Security Level
of resources in processing high-quality image frames every
Every organization needs to secure their premises for the time which is actually not required. We need to process all
unknown entry into that place. They also wish to monitor the data and wants to store all the data creates a huge space.
the employees and indusial entry into that place. Those who But using face recoginization we can done this job in
are entering the organization premises without proper access fraction of seconds using artificial intelligence concept. To
they are capturing in the security surveillance system and reduce the speed of processing photo frames professional
notice to the respective person and alerts instantly agencies will use clusters of computers. But each additional
concerning the person who doesn’t have permission.[11] system means that appreciable information transfer via
B. Straightforward Integration method internet creates a problem.
The automatic face detection tools works effectively with B. Size & Quality
the current authentication code that organizations has To operate face recognition system correctly and perfectly
developed.Bascially the technique is a straightforward to we required advance software system using high quality
code the system to access organizations automatic data digital cameras.identfication system takes a photo of a
processing which makes the method very clear. person or takes a screen shot from the video and starts
C. High Accuracy Rates comparing with the actual image hence here the storage
The main advantage is its Accuracy. The system checks and really matters and affects the storage.ess the image to reduce
gives the output without any misunderstanding and bad face the size of the image it will affects the quality of the system
detection system. The authorized person will be detected at in recognizing the face. For example consider CCTV and
the right time due to the high accuracy levels. The manual the person who is far away the camera recognizes the person
recognition, which is done by securities outside of the and takes photo from a long distance if the resolution of the
organization’s premises we may use the face recognition image is decrees then we can’t identify the person if we
technology to automate the process of identification and didn’t decreases the quality of the image the storage size of
assures its perfection without changes. We don’t want the image is very high.
addition employee to monitor the working of cameras 24/7.
The main objective of Automation means to reduce human
Published By:
Retrieval Number: B10571292S19/2019©BEIESP Blue Eyes Intelligence Engineering
DOI: 10.35940/ijitee.B1057.1292S19 595 & Sciences Publication
International Journal of Innovative Technology and Exploring Engineering (IJITEE)
ISSN: 2278-307A5, Volume-9 Issue-2S, December 2019
To avoid bad detection of images and speedup the It is a biometric system which uses the Iris-Recognition
recoginization process we need to permit the identification management .this system uses the concept of iris
in face-size range. The main problem with the system is recoginization system in which the iris are scanned and
initial investment in such a huge software system is very extracted the features and matched with the database. The
expensive and the processing speed will decreases due to main problem with this system is placing the transmission
high quality of the image. lines of the scanner in a good condition of light to scan the
entire irises passing through the system. It is based on the
C. Surveillance Angle
real time face detection system which is highly secure,
The police identification method is additionally creates reliable and fast to access but needs a lot of development in
problems in many ways which was answerable for the different lighting conditions was discussed by Akshara
selected face capturing by the camera to register a face using Jadhav et al[1]
software recognition system, so many angles are getting
used like 45 degree, frontal face etc. To get transparent VIII. CONCLUSION:
model for the image, we have to use frontal part. One can
We are automating the attendance system to decrease the
fool the face recoginization system by hair on face,
errors occurs due to the manual taking attendance. If the
spectacles etc but using the high intention image with direct
cameras monitoring into classrooms to evaluate their interest
angle goes with the enrolled and compared with the actual
and to mark attendance, students tends to pay attention if
image to get accurate results. Even a person can fooli system
Artificial Intelligence enabled method can monitor and
by appearing suddenly or removing beard or mask on face
mark their attendance and faculties will at least come to
etc. We can simply avoid the above mentioned problems by
school or college every day because, in early times they are
updating the databases regular To avoid such failures, the
coming and putting sign and they are letting the school or
databases must be regularly updated with all the recent
college now it’s not possible if the faculty left the college
changes.
the system automatically marks as absent so everyone will
come to school or organization regularly. Using artificial
VII. RELATED WORKS:
intelligence concept the attendance monitoring system is
A. Fingerprint-based attendance System very secure, accurate and easy to monitor students and
This attendance system works on the moveable fingerprint faculty’s attendance.[8]
device developed and can pass the device to all students to
scan their thump impression on the device during the class REFERENCES:
time to mark attendance without faculty involvement. This [1] Akshara Jadhav, Akshay Jadhav Tushar Ladhe, Krishna Yeolekar
method assures the never failure method for taking the “Automated Attendance System Using Face Recognition” Irjet
Volume: 04 Issue: 01 | Jan -2017
attendance. The main issue with this method is passing the [2] Lerato Masupha,Tranos Zuva,Seleman Ngwira,Omobayo Esan “Face
fingerprint device to all the students in class time that may Recognition Techniques, Their Advantages, Disadvantages And
Performance Evaluation” 2015 International Conference On
distrust the students concentration which is of no use for the Computing, Communication And Security(Icccs),1-5,2015.
students and difficult to pass the device without disturbing [3] Jireh Robert Jam, Face Detection And Recognition Student
the class to all students to mark attendance by putting their Attendance System,Kingston University London
[4] Expert Opinion: The Effect Of Artificial Intelligence On Face
thumb on the device was discussed by Akshara Jadhav et al Recognition Avaliable At Https://Www.Everteam.Com/En/Artificial-
[1]. Intelligence-Face-Recognition/ Accessed On (08/10/2019)
[5] Pradeepa .M, H P Mohan Kumar” Face Detection And Recognition
B. Radio Frequency Identification-based Attendance For Automatic Attendance System Using Artificial Intelligence
Concept”Ijesc Volume 8 Issue No.5 May 2018
System [6] Santana Fell “Ai To Mark Attendance” The Telegraph Online Edition
Published On Wednesday, 25 September 2019
The works connected to this identification attendance [7] Chaitanya Reddy “Face Recognition Using Artificial Intelligence”
system still present in the practical society described by the Towardsdatascience.Com In April 2017.
author Akshara Jadhav at el. they proposed an Radio [8] Sulakshana Iyer “Facial Recognition Technology Is Too Great To
Ignore” On Soft Web Solutions In March 6th, 2019
frequency identification system. The working is this type [9] Shreyak Sawhney ; Karan Kacker ;Samyak Jain ;Shailendra Narayan
system every student needs to carry a ID card called radio Singh ;Rakesh Garg “Real-Time Smart Attendance System Using
Face Recognition Techniques” Published In Ieee 2019 9th
frequency tag and wants to put the ID card on card reader to International Conference On Cloud Computing, Data Science &
read and to store the attendance of the student in database. Engineering (Doi: 10.1109/Confluence.2019.8776934)
[10] Jyothi S Bedre,Shubhangi Sapkal “Comparative Study Of Face
The main problem with this techniques is unauthorized Recognition:A Review” In Emerging Trends In Computer Science
access. The person using other person id card can enter into And Information Technology-2012 And Precedings Published In
International Journal Of Computer Applications.
organization which the process is less secure and even they [11] Refik Samet,Muhammed Tanriverdi “Face Recognition-Based Mobile
mark attendance of their friends by approved ID card Automatic Classroom Attendance Management System” Published In
scanning on the scanner. Ieee 2017 International Conference On Cyberworlds (Cw)
(Doi: 10.1109/Cw.2017.34)
C. Iris-Recognition Based Attendance System
Published By:
Retrieval Number: B10571292S19/2019©BEIESP Blue Eyes Intelligence Engineering
DOI: 10.35940/ijitee.B1057.1292S19 596 & Sciences Publication
AI-Based Attendance Monitoring System
AUTHORS PROFILES
K P Naveen Reddy- obtained his B. Tech degree
(CS&E Branch) in GITAM School of technology,
Bangalore campus. He has published the research
paper titled “A study of Robotic Process
Automation Among Artificial Intelligence” on
International Journals of science and research
publications, Volume 9, Issue 2, February 2019
ISSN 2250-3153 and two more article with name
"Closed Form Solutions Of Funcanality of Windshield Wipers" in IJEE
(Volume 10, Issue 1) and “Comparison of Programming Languages:
Review” in IJCSE (Volume 9 Issue 2) which are UGC Recommended
Journals. The research aims are Artificial intelligence, Robotics, Machine
learning and Big data.
Published By:
Retrieval Number: B10571292S19/2019©BEIESP Blue Eyes Intelligence Engineering
DOI: 10.35940/ijitee.B1057.1292S19 597 & Sciences Publication
View publication stats