This document summarizes a proposed attendance tracking system using facial recognition. The system allows instructors to take attendance automatically by taking a photo of the class with their phone, or manually enter attendance on a web or mobile app. Data is stored both online and can be synced offline. Reports can be exported to Excel. The teaching calendar can also be synced with Google Calendar. The current accuracy of the facial recognition is 70-80% and it only works for classrooms of 20-30 students. Future plans include improving accuracy and supporting more platforms/report types.

1. Group 1

2. Team Member
Supervisor: Kiều Trọng Khánh
Phạm Huy Hoàng
Nguyễn Thanh Bình
Nguyễn Quốc Huy
Đỗ Minh Đạt (drop out)

3. Outline
Scenario Problem
Advantages & Disadvantages of current scenario
Propose our solution
Demostration
Techology and development evironment
Advantages
Disadvantage & Future Plan

4. Scenario Problem
 The demand of attendance taking.
 Traditional attendance taking method has many
disadvantages.

5. Traditional Method
Instructor 1
Instructor 2
Student 1
Student 2
Student 3
Student 4
Roll Call
Book 1
Information
and Report
Roll Call
Book 2
1
1
1
1
2
2
2
2
3
3

6. Advantages Disavantage

7. Fingerprint Recognition Method
Student 1
Student 2
Fingerprint
Recognition
Device 1
Student 3
Student 4
Fingerprint
Recognition
Device 2
Information
and Report
Instructor
Storage
Browser
1
1
1
1
2
2
3 4
5
CMS

8. Advantages Disavantage

9. Proposed scenario
Students
Information
and Report
Instructor
Storage
Browser
Instructor
Cellphone
Web Services
Staff
Student
1
2
3
4
5
6
7 8
9

10. Features
Check Attendance Face Recognition Export Excel Report
Mobile Application Google Calendar Sync Offline Data Sync

11. Demo

12. Demo scenario
Display Student
Images and current
Roll Calls in System
Check Attendance by
Taking Picture
Check Attendance
Manually
(Web & Mobile)
Check Attendance
Offline & Sync Data
Online
Export Excel Report
Manage Study Session.
Sync Calendar to
Google Calendar

13. How Face Recogniton work
FACE RECOGNIZER
TRAINED
Quang
Huy
Thanh
Bình

14. Demo scenario
Display Student
Images and current
Roll Calls in System
Check Attendance by
Taking Picture
Check Attendance
Manually
(Web & Mobile)
Check Attendance
Offline & Sync Data
Online
Export Excel Report
Manage Study Session.
Sync Calendar to
Google Calendar

15. Check Attendance Auto
Instructor
Students
Web Services
Attendance
Checking
Recognize
Result
1
2
3
Face Recognition
4
5
5
6
7

16. Demo scenario
Display Student
Images and current
Roll Calls in System
Check Attendance by
Taking Picture
Check Attendance
Manually
(Web & Mobile)
Check Attendance
Offline & Sync Data
Online
Export Excel Report
Manage Study Session.
Sync Calendar to
Google Calendar

17. Check Attendance Manual & Offline
4
1
2
3
5
6
1
2
3
4
5
Flow sql server
Flow sqlite
Using website

18. Demo scenario
Display Student
Images and current
Roll Calls in System
Check Attendance by
Taking Picture
Check Attendance
Manually
(Web & Mobile)
Check Attendance
Offline & Sync Data
Online
Export Excel Report
Manage Study Session.
Sync Calendar to
Google Calendar

19. Excel Report
Instructor
Staff
Roll Call Book
Student Attendance
Report
Student List for
Final Exam

20. Demo scenario
Display Student
Images and current
Roll Calls in System
Check Attendance by
Taking Picture
Check Attendance
Manually
(Web & Mobile)
Check Attendance
Offline & Sync Data
Online
Export Excel Report
Manage Study
Session.
Sync Calendar to
Google Calendar

21. Sync Teaching Calendar
Instructor
Website
System
Google Service
Teaching
Calendar
Google API
Google
Calendar
View Teaching
Calendar
Sync Teaching
Calendar
1
2
3
1
2
3
4 5
6

22. Achievement
 Fast, convinient roll call system. The attendance
checking only take 10-30s for each session.
 Small installation cost.
 Easy deployment (Just web server + mobile).
 Less hand work.
 Offline working and data synchronization, in case of
network error.

23. Disadvantage
 The accuracy is still not 100%. (Just 70-80% according
to test report).
 Only work with small class room/examine room (From
20~30 students).
 Only export one kind of report.
 Mobile App only on 1 platform: Android

24. Future Plan
 Find another face-recognition algorithm or optimize
the current algorithm for higher accuracy.
 Export more kind of report: Absent rate of a subject, a
student.
 Develop mobile app on another platform: iOS,
Window Phone…

25. THANK YOU
Q/A