DNN BASED SMART ATTENDANCE

MANAGEMENT SYSTEM

existing framework with the assistance of machine learning,

Abstract— To keep up the participation record with it requires a decent quality camera to catch the pictures of
everyday exercises is a difficult undertaking. The
accompanying framework depends on face acknowledgment to
understudies, the identification cycle is finished by histogram
keep up the participation record of understudies. The every of arranged slope. What's more, perceiving perform through
day participation of understudies is recorded subject astute profound learning. The frontend side (customer side) which
which is put away as of now by the head. As the ideal comprise of GUI which depends on electron JS and backend
opportunity for relating subject shows up the framework side comprise of rationale and MATLAB (worker side), an
naturally begins taking snaps and afterward apply face IPC (Bury Individual Correspondence) connect is created to
location and acknowledgment procedure to the given picture
and the perceive understudies are set apart as present and
convey these two stacks. The pictures catch by the camera is
their participation update with comparing time and subject id. shipped off framework for additional investigation, the
We have utilized Deep learning strategies to build up this information picture is then contrasted and a bunch of
framework, Histogram of Oriented Gradients strategy, Harris reference pictures of every one of the under study and
Corner Locator is utilized to distinguish faces in pictures and imprint their participation.
Profound learning technique is utilized to process and contrast
include facial of understudies with remember them. Our
framework is competent to distinguish different appearances Related Work
progressively. Lately, various face acknowledgment based participation
the executives framework have presented all together to
Keywords: Facial Recognition, Deep Learning, HOG, Harris
improve the exhibition of understudies in various
Corner.
association. In [4] Jomon Joseph, K. P. Zacharia proposed a
I. INTRODUCTION framework utilizing picture preparing, PCA, Eigen faces,
Microcontroller, in view of Matlab. Their framework works
Face Each association requires a hearty and stable just with front face pictures and there is need of a reasonable
framework to record the participation of their understudies. technique which works with the direction of the framework.
what's more, every association have their own strategy to do Ajinkya Patil with their colleagues in [5] proposed a face
as such, some are gauging participation physically with a acknowledgment approach for participation checking
piece of paper by calling their names during address hours utilizing Viola jones calculation, Haar falls are utilized to
and some have received biometrics framework like unique recognize faces in pictures and acknowledgment performs
mark, RFID card per user, Iris framework to stamp the through Eigen face technique. Another methodology of
participation. The ordinary technique for calling the names making participation framework simple and secure, in [6]
of understudies physically is tedious occasion. The RFID the creator proposed a framework with the assistance of
card framework, every understudy allots a card with their counterfeit neural organizations, they utilized PCA to
comparing personality however there is possibility of card extricate face pictures and testing and preparing were
misfortune or unapproved individual may abuse the card for accomplished by neural organizations, their framework
counterfeit participation. While in other biometrics, for performs in different direction. A 3D face acknowledgment
example, unique finger impression, iris or voice approach for participation the executives framework was
acknowledgment, they all remain imperfect and additionally proposed by MuthuKalyani.K, VeeraMuthu.A [7] has
they are not 100% exact [1] [19]. Utilization of face proposed, they checked participation with month to month
acknowledgment with the end goal of participation stamping progress of every understudy. There is need for an elective
is the keen method of participation the executives calculation which can upgrade the acknowledgment on
framework. Face acknowledgment is more precise and situated countenances. Proficient Participation The board
quicker method among different procedures and lessens framework is planned with the assistance of PCA
possibility of intermediary participation. Face calculation [8], the have accomplished precision up to 83%
acknowledgment give detached ID that is an individual yet their framework execution diminishes due to somewhat
which is to be recognized doesn't to have to make any move changes in light condition. An eigen face approach
for its personality [2]. Face acknowledgment includes two alongside PCA calculation for stamping face
stages, initial step includes the recognition of countenances acknowledgment participation framework have presented by
and second step comprise of distinguishing proof of those creator in [9], they notice correlation of various face
recognized face pictures with the existing data set. There are acknowledgment calculation in their paper. Generally
number of face identification and acknowledgment speaking it was acceptable way to deal with keep up record
techniques presented. Face acknowledgment works either in of participation..
type of appearance based which covers the highlights of
entire face or highlight based which covers the mathematical
element like eyes, nose, eye foreheads, and cheeks to II. SYSTEM OVERVIEW
perceive the face [3].Our framework utilizes face
acknowledgment way to deal with lessen the blemishes of The proposed framework is intended for computerizing
the participation of the distinctive association and lessens
1
 the imperfections of existing manual framework. The
framework ascertain the participation subject astute, that is furthermore, indistinct information structure. On the off chance
the information of understudies and subjects are added that you are attempting to store or trade information in
physically by director, and at whatever point time for capacities or dates than JSON isn't right decision for you.
comparing subject shows up the framework consequently
begins taking snaps and discover whether human
countenances are show up in the given picture or not. We 3.2 Face Recognition Process
have utilized Histogram of Arranged Inclination for face
identification and profound learning procedures to figure
and analyze 128-d face highlights for face
acknowledgment. Once faces are distinguished and
perceive with the current data set, framework figure
participation for the perceive understudies with the separate
subject id continuously. Furthermore, an dominate sheet
created and saved by the framework naturally. Our
framework parts into two sections, First the front end side
which comprise of GUI which depends on Electron JS that
is JavaScript stack which is filling in as a customer and the
second is the backend side which comprise of rationale and 3.2.1 Face Identification and Extraction: Face recognition
in light of MATLAB which is filling in as a worker. Also, is significant as the picture taken through the camera given to
we realize that both the dialects can't speak with each other the framework, face discovery calculation applies to recognize
straightforwardly so we have utilized IPC (Entomb the human faces in that picture, the quantity of picture handling
Individual Correspondence) strategies with zero library as a calculations are acquaint with distinguish faces in a pictures and
scaffold to convey these two dialects. The Electron JS call furthermore the area of that distinguished countenances. We
the MATLAB capacities and exchange information through have utilized Hoard strategy to recognize human countenances
TCP with help of Zero PC Library in given picture.

3.2.2 Face Positioning: There are 68 explicit focuses in a

human face. All in all we can say 68 face tourist spots. The
3.1 Data Acquisition
principle capacity of this progression is to identify tourist spots
3.1.1 Image obtaining: Picture is get utilizing a top quality
of appearances and to situate the picture. A MATLAB content
camera which is put in the study hall. This picture is given as a
is utilized to consequently identify the face tourist spots and to
contribution to the framework.
situate the face however much as could reasonably be expected
without mutilating the picture.
3.1.2 Dataset Creation: Dataset of understudies is made
before the acknowledgment interaction. Dataset was made
3.2.3 Face Encoding: When the appearances are identified
distinctly to prepare this framework. We have made a dataset
in the given picture, the subsequent stage is to extricate the
of 5 understudies which includes their name, move number,
remarkable distinguishing facial element for each picture.
division and pictures of understudy in various stances and
Fundamentally at whatever point we get restriction of face, the
varieties. For better precision least 15 pictures of every
128 key facial point are separated for each picture given info
understudies ought to be caught. At whatever point we register
which are profoundly exact and these 128-d facial focuses are
understudy's information and pictures in our framework to
put away in information document for face acknowledgment.
make dataset, profound learning applies to each face to figure
128-d facial highlights and store in understudy face
3.2.4 Face coordinating: This is last advance of face
information record to review that face in acknowledgment
acknowledgment measure. We have utilized the a standout
measure. This interaction is applies to each picture taken
amongst other learning procedure that is profound measurement
during enrollment.
realizing which is exceptionally exact and equipped for yielding
genuine worth component vector. Our framework confirms the
3.1.3 Storing: We have utilized JSON to store the
countenances, building the 128-d inserting (confirmation) for
understudy's information.
each. Inside compare_faces work is utilized to register the
Euclidean distance between face in picture and all appearances
JavaScript Object Notation(JSON): To address an organized
in the dataset. In the event that the current picture is coordinated
information dependent on JavaScript object language structure,
with the 60% limit with the current dataset, it will move to
a standard book based organization is presented. JSON is
participation.
utilized for sending information in web application. It is an
ideal answer for putting away impermanent information that is
devoured by the element that is makes the information. JSON
can store information in String, Number, Article, Cluster,
Boolean, Invalid structure which implies it has impediment of
putting away information in capacities, dates 3.3 Attendance Marking

Keen Participation The executives Framework Utilizing Face When the face is relate to the picture put away in JSON
Acknowledgment document, python create move quantities of present
2
understudies and return that, when information is returned, the
framework produces participation table which incorporates the
name, move number, date, day and time with comparing
subject id. And afterward passes the information to python to
store the table into a dominate sheet consequently. Each sheet
is saved by the subjects which previously entered by the
chairman, for instance when framework produces dominate
sheet by sending the arranged sheet in an exhibit to python, the
python first checks whether there leave any dominate sheet of
that date, assuming indeed, it make separate worksheet by
subject id, so participation is separated for various subjects.

4. Results and Analyses • Addition of subject with their relating time.

Smart Attendance Management System is simple and works • Attendance sheet age and import to Excel (xlsx)
efficiently.The board Framework is basic and works format.
effectiveness. The framework works naturally once the To instate this framework, the manager first register their
enlistment of individual understudy made by the organization understudy information alongside their name move number and
office. We have made a preparation dataset of 6 understudies
This is first page of our participation framework. Which (absolute of 120 pictures for each) for testing reason.
depends on HTML5, CSS3 and JS. It comprise of the
accompanying modules, This is the subject envelope, subjects are to be occupied by time
table once the time shows up for the relating subject, the
• Student Enrollment framework begins catching pictures, distinguishes the faces,
contrasts the countenances and existing data set, mark
• Face Acknowledgment participation and create dominate sheet for the perceive
understudies.

III. CONCLUSION
The Smart attendance management system is intended to
address the issues of existing manual frameworks. We have
utilized face acknowledgment idea to check the participation
of understudy and improve the framework. The framework
performs palatable in various stances and varieties. In future
this framework need be improved in light of the fact that
these framework at times neglects to perceive understudies
from some distance, likewise we make them measure
impediment, working with an arrangement of high preparing
may result surprisingly better execution of this framework.

Face Orientations Detection Rate Recognition Rate acknowledgment innovation in the

foundations/associations. The framework gauges
O0 (Frontal face) 98.7 % 95% participation of every understudy by constant perception at
the passage and leave focuses. The consequence of our
starter analyze shows improved execution in the assessment
18º 80.0 % 78%
of the participation contrasted with the customary high
54º 59.2 % 58% contrast participation frameworks. Current work is centered
72º 0.00 % 0.00% around the face discovery calculations from pictures or
90º(Profile face) 0.00 % 0.00% video outlines.

In additional work, creators plan to improve face

acknowledgment viability by utilizing the communication
among our framework, the clients and the heads. Then
IV CONCLUSION AND FUTURE WORK again, our framework can be utilized in a totally new
element of face acknowledgment application, versatile
To acquire the participation of people and to record their
based face acknowledgment, which can be a guide for
season of passage and leave, the creators proposed the everyday citizens to think about any individual being shot
participation the executives framework dependent on face
3
 by cell camera including legitimate approval for getting to [17] Karnalim, Oscar, et al. "Face-face at classroom environment:
a unified information base. Dataset and exploration." 2018 Eighth International
Conference on Image Processing Theory, Tools and
Applications (IPTA). IEEE, 2018.
[18] Mian, Ajmal. "Realtime face detection and tracking using a
single pan, tilt, zoom camera." 2008 23rd International
Conference Image and Vision Computing New Zealand.
REFERENCES IEEE, 2008.
[19] Mehta, Preeti, and Pankaj Tomar. "An Efficient Attendance
[1] Kar, Nirmalya, et al. "Study of implementing automated Management Sytem based on Face Recognition using
attendance system using face recognition technique." Matlab and Raspberry Pi 2." International Journal
International Journal of computer and communication of Engineering Technology Science and Research IJETSR
engineering 1.2 (2012): 100. 3.5 (2016): 71-78.
[2] RoshanTharanga, J. G., et al. "Smart attendance using real
time face recognition (smart-fr)." Department of Electronic
and Computer Engineering, Sri Lanka Institute of
Information Technology (SLIIT), Malabe, Sri Lanka (2013)
[3] Selvi, K. Senthamil, P. Chitrakala, and A. Antony Jenitha.
"Face recognition based attendance marking system."
Corresponding Author: S. Rajkumar*, Email:
rajkumarsrajkumar@ gamil. com (2014).
[4] Joseph, Jomon, and K. P. Zacharia. "Automatic attendance
management system using face recognition." International
Journal of Science and Research (IJSR) 2.11 (2013): 327-
330.
[5] Patil, Ajinkya, and Mrudang Shukla. "Implementation of
classroom attendance system based on face recognition in
class." International Journal of Advances in Engineering &
Technology 7.3 (2014): 974.
[6] Kanti, Jyotshana, and Shubha Sharm. "Automated
Attendance using Face Recognition based on PCA with
Artificial Neural Network." International journal of
science and research IJSR(2012).
[7] MuthuKalyani, K., and A. VeeraMuthu. "Smart application
for AMS using face recognition." Computer Science &
Engineering 3.5 (2013): 13.
[8] Deshmukh, Badal J., and Sudhir M. Kharad. "Efficient
Attendance Management: A Face Recognition Approach."
(2014).
[9] Wagh, Priyanka, et al. "Attendance system based on face
recognition using eigen face and PCA algorithms." 2015
International Conference on Green Computing and
Internet of Things (ICGCIoT). IEEE, 2015.
[10] Bhattacharya, Shubhobrata, et al. "Smart Attendance
Monitoring System (SAMS): A Face Recognition Based
Attendance System for Classroom Environment." 2018
IEEE 18th International Conference on Advanced
Learning Technologies (ICALT). IEEE, 2018.
[11] Samet, Refik, and Muhammed Tanriverdi. "Face
recognition-based mobile automatic classroom attendance
management system." 2017 International Conference on
Cyberworlds (CW). IEEE, 2017.
[12] Li, Xiang-Yu, and Zhen-Xian Lin. "Face recognition based
on HOG and fast PCA algorithm." The Euro-China
Conference on Intelligent Data Analysis and Applications.
Springer, Cham, 2017.
[13] Arsenovic, Marko, et al. "FaceTime—Deep learning based
face recognition attendance system." 2017 IEEE 15th
International Symposium on Intelligent Systems and
Informatics (SISY). IEEE, 2017.
[14] Rekha, N., and M. Z. Kurian. "Face detection in real time
based on HOG." International Journal of Advanced
Research in Computer Engineering & Technology
(IJARCET) 3.4 (2014): 1345-1352.
[15] Kwolek, Bogdan. "Face detection using convolutional
neural networks and Gabor filters." International
Conference on Artificial Neural Networks. Springer,
Berlin, Heidelberg, 2005.
[16] Ashwini, C., et al. "An Efficient Attendance System Using
Local Binary Pattern and Local Directional
Pattern." Journal of Network Communications and
Emerging Technologies (JNCET) www. jncet. org 8.4
(2018).
4
5