Smart Class Room Attendance System using Group Face

Recognition
Objective
The basic aim is to automate the system and implement the smart class room which is useful for
educational organizations.
The followings are the objectives of this project:
 To develop a portable Smart Attendance System which is handy and self-powered?
 To ensure the speed of the attendance recording process is faster than the previous system
which can go as fast as approximately 3 second for each student.
 Have sufficient memory space to store the database.
 Able to recognize the face of an individual accurately based on the face database.
 Allow parents to track their child’s attendance.
 Develop a database for the attendance management system.
 Provide a user friendly web interface for admins to access the attendance database and for
non-admins (parents) to check their child’s attendance.
 Allow new students or staff to store their faces in the database by using a GUI.
 Able to show an indication to the user whether the face- recognition process is successful
or not.

Abstract
The human face is an important entity which plays a crucial role in our daily social interaction,
like conveying individual's identity. Face recognition system is also able to recognize the person
from a distance without touching or any interaction with the person. Currently, face recognition
applications are deployed in social media websites like Facebook, in the entrance of Airports,
Railways Stations, Bus Stop, highly secured areas, advertisement, and health care. The purpose
of these applications is to minimize criminal activities, fake authentication, tracking addictive
gamblers in casinos, whereas Facebook is using face recognition system for automatic tagging
purpose. For face recognition purpose, there is a need for large data sets and complex features to
uniquely identify the different subjects by manipulating different obstacles like illumination,
pose and aging.in this project we proposes a deep unified model for Face Recognition based on
Faster Region Convolution Neural Network. Design a group-based face attendance system based
on the proposed deep unified model. In our proposed system, we have a number of class rooms
of a specific institute in which we setup our face recognition system for making a smart class
rooms. Several images from different smart class room Buffys are being sent simultaneously for
processing, in order to take the attendance. In order to measure the validity of the proposed
algorithm, a web application of a groupbased face attendance system is developed.

Software’s and Hardware’s Requirement

Recommended System Requirements

 Processors:
 Intel® Core™ i5 processor 4300M at 2.60 GHz or 2.59 GHz (1 socket, 2 cores, 2
threads per core), 8 GB of DRAM
 Intel® Xeon® processor E5-2698 v3 at 2.30 GHz (2 sockets, 16 cores each, 1 thread
per core), 64 GB of DRAM
 Intel® Xeon Phi™ processor 7210 at 1.30 GHz (1 socket, 64 cores, 4 threads per
core), 32 GB of DRAM, 16 GB of MCDRAM (flat mode enabled)
 Disk space: 2 to 3 GB
 Operating systems: Windows® 10, macOS*, and Linux*
Minimum System Requirements

 Processors: Intel Atom® processor or Intel® Core™ i3 processor

 Disk space: 1 GB
 Operating systems: Windows* 7 or later, macOS, and Linux
Software Specification

 Python* versions: 2.7.X, 3.6.X

 Included development tools: conda*, conda-env, Jupyter Notebook* (IPython)
 Compatible tools: Microsoft Visual Studio*, PyCharm*
 Included Python packages: NumPy, SciPy, scikit-learn*, pandas, Matplotlib, Numba*,
Intel® Threading Building Blocks, pyDAAL, Jupyter, mpi4py, PIP*, and others.
  PIP and NumPy: Installed with PIP, Ubuntu*, Python 3.6.2, NumPy 1.13.1, scikit-learn
0.18.2
 Windows: Python 3.6.2, PIP and NumPy 1.13.1, scikit-learn 0.18.2
 Intel® Distribution for Python* 2018
 TkInter Editor for Python

Hardware’s
 Webcamera
 Laptop or PC

Existing System
 Fingerprint Based recognition system: In the Fingerprint based existing attendance
system, a portable fingerprint device need to be configured with the students fingerprint
earlier. Later either during the lecture hours or before, the student needs to record the
fingerprint on the configured device to ensure their attendance for the day. The problem
with this approach is that during the lecture time it may distract the attention of the
students.
 RFID(Radio Frequency Identification) Based recognition system: In the RFID based
existing system, the student needs to carry a Radio Frequency Identity Card with them
and place the ID on the card reader to record their presence for the day. The system is
capable of to connect to RS232 and record the attendance to the saved database. There
are possibilities for the fraudulent access may occur. Some are students may make use of
other students ID to ensure their presence when the particular student is absent or they
even try to misuse it sometimes.
 Iris Based Recognition System: In the Iris based student attendance system, the student
needs to stand in front of a camera, so that the camera will scan the Iris of the student.
The scanned iris is matched with data of student stored in the database and the attendance
on their presence needs be updated. This reduces the paper and pen workload of the
faculty member of the institute. This also reduces the chances of proxies in the class, and
 helps in maintaining the student records safe. It is a wireless biometric technique that
solves the problem of spurious attendance and the trouble of laying the corresponding
network.
 Face Based Recognition System: The facial recognition technology can be used in
recording the attendance through a high-resolution digital camera that detects and
recognizes the faces of the students and the machine compares the recognized face with
students’ face images stored in the database. Once the face of the student is matched with
the stored image, then the attendance is marked in attendance database for further
calculation. If the captured image doesn't match with the students' face present in the
database then this image is stored as a new image onto the database. In this system, there
are possibilities for the camera to not to capture the image properly or it may miss some
of the students from capturing.
Existing Algorithms for face recognition System

 PCA
 LDA
 HOG
 K-nn
 SVM

Disadvantages

 The accuracy of the system is not 100%.

 Face detection and loading training data processes just a little bit slow.
 It can only detect face from a limited distance.
 It cannot repeat live video to recognize missed faces.

 The instructor and trainingSet manager still have to do some work manually.

Proposed System
 Propose a deep unified model for Face Recognition based on Faster Region Convolution
Neural Network.
  Proposes an algorithm for face detection and recognition based on deep Convolution
Neural Networks (CNN).
 Our proposed system not only detects the appropriate number of faces from the frame,
but also recognizes the detected faces.
 Face detection purpose, Region Proposal Network (RPN) draws anchors and outputs the
one which most likely contains the objects.
 Deep convolution neural network architecture is developed for the recognition.
 The edge computing processes the data at the edges of the nodes, here edge is a
computing device and network resource along with the dedicated path of generated data
sources and cloud data centers.
Advantages

 The system stores the faces that are detected and automatically marks attendance.
 Provide authorized access.
 Ease of use.
 Multiple face detection.
 Provide methods to maximize the number of extracted faces from an image.
 The system stores the faces that are detected and automatically marks attendance.
 Ease of use.
 Manipulate and recognize the faces in real time using live video data.
 Multiple face detection.
 Multipurpose software.
 Can be used in different places.

Concept Diagram
Architecture Diagram

Enrollment Attendance

Face Data Set acquisition

Preprocessing

Preprocessing
Face Detection

Face Detection
Face Recognition

Face Recognition
Feature Extraction

Feature Extraction
Classification

Classification

Classified Result

Student
Database Matching
Server
College Admin

Staff

Parent

Modules
 1. Enrollment Module
1.1. Training
2. Verification Module
2.1. Testing
3. Attendance System
4. Notification
5. Performance Analysis

Problem Description
In our proposed system, we have a number of class rooms of a specific institute in which we set
up our face recognition system for making a smart class rooms. Several images from different
smart class room Buffys are being sent simultaneously for processing, in order to take the
attendance. All class rooms are connected to the gateway device which is placed in some central
place of the institute. At the time of the registration of the user, all the data is passed to the
clouds, because at that time data latency rate and response time does not matter. The necessary
cloud data is synchronized into the gateway device after the particular time stamp. But at the
time when teacher upload the attendance quick response time really matters. To getting the
names of the recognized persons, at the time of attendance the captured image from the device is
passed to the gateway which is edge node in our case. The predicted names are passed back to
the user interface, from which the image is captured. After taking the decisions based on the
prediction the names are passed to the cloud for generating reports of the attendance of the
specific day. The system flow of this process is shown in Figure 2. Based on the prediction the
system generates, the teacher or smart class room will have to take actions, like door lock, turn
on the electronic white boards etc, so this process must be a performed in an efficient manner. In
our proposed edge computing architecture, at the time of attendance, the uploaded image is
passed to the gateway device, it returns the predicted information in a shorter time per
comparison to the time it would take if the data sent to the cloud. The architecture includes the
User Interface and Backend services developed in python.

Modules Description
1. Enrollment Phase
Student Databases Server maintained in this system are student information database, face
database and attendance database. Student information database consists of roll number, name
and class of student. Attendance database includes attendance status of student for every day.
The face database consists of face images of student’s according to their roll numbers.
1.1. Face Image Acquisition-
This module is initial part of the system. Logitech C270 (3MP) is used for image acquisition.
1.2. Preprocessing
The acquired images are converted to grayscale image and then resize. After the removal of
noise using mean and Gaussian filters all further operations are performed on this image.
1.3. Face Detection
After capturing the image, the image is given to face detection module. This module detects the
image regions which are likely to be human. Face detection purpose, Region Proposal Network
(RPN) draws anchors and outputs the one which most likely contains the objects. The detected
face regions are cropped and scaled to 200x200 resolution and then used for further recognition
task.
1.4. Feature Extraction
After the face detection, face image is given as input to the feature extraction module to find the
key features that will be used for classification.
1.5. Feature Classification
The module composes a very short feature vector that is well enough to represent the face image.
Here, it is done with DCNN method. Then the classified result is stored into the database.

2. Verification Module
After capturing the image, the image is given to face detection module. This module detects the
image regions which are likely to be human. After the face detection using RPN, face image is
given as input to the feature extraction module to find the key features that will be used for
classification. The module composes a very short feature vector that is well enough to represent
the face image. Here, it is done with DCNN with the help of a pattern classifier, the extracted
features of face image are compared with the ones stored in the face database. The face image is
then classified as either known or unknown. If the image face is known, corresponding student is
identified. Face recognition is the next step after face detection. The faces cropped from the
image are compared with the stored images in the face database. Here, Eigen face method is used
for face recognition.

3. Attendance System
After the verification of faces and successful recognition is done, the attendance of the student is
marked in front of his/her roll number. If the faceis not recognized, an error page is displayed. It
involves the attendance report generation. The module takes student information and daily
attendance status from student database. The attendance is calculated as per requirement. There
are options for calculating day-wise, student wise and class-wise attendance. The attendance
reports are generated and saved in a file.

4. Notification Services
Our students’ parents have access the notification services, where all the marks and activities, as
well as the student’s attendance, are recorded; now, they will also be notified of the attendance
by SMS. The parent knows where their child is at any time – and this is one of the greatest
parental cares and interests. Even when you have no internet access, you receive messages on
your mobile device, notifying you of your child's arrival time and any missed lessons, so you can
check on your child right away and provide assistance if needed.
5. Performance Comparison
There are various parameters with the help of which one can measure the performance of any
biometric authentication techniques. These factors are described below.
False Accept Rate (FAR) or False Match Rate (FMR): The probability that the system
incorrectly declares a successful match between the input pattern and a non matching pattern in
the database. It measures the percent of invalid matches. These systems are critical since they are
commonly used to forbid certain actions by disallowed people.
False Reject Rate (FRR) or False Non-Match Rate (FNMR): The probability that the system
incorrectly declares failure of match between the input pattern and the matching template in the
database. It measures the percent of valid inputs being rejected.
Equal Error Rate (EER): The rates at which both accept and reject errors are equal. ROC or
DET plotting is used because how FAR and FRR can be changed, is shown clearly. When quick
comparison of two systems is required, the ERR is commonly used. Obtained from the ROC plot
by taking the point where FAR and FRR have the same value. The lower the EER, the more
accurate the system is considered to be.
Failure to Enroll Rate (FTE or FER): The percentage of data input is considered invalid and
fails to input into the system. Failure to enroll happens when the data obtained by the sensor are
considered invalid or of poor quality.
Failure to Capture Rate (FTC): Within automatic systems, the probability that the system fails
to detect a biometric characteristic when presented correctly is generally treated as FTC.