Vision Based Automated Attendance System Develop Using Deep Learning Techniques
A-SYS is an android based mobile application for marking the attendance of students sitting in class. Attendance is marked by detecting the faces of students and the picture is captured through the instructor?s mobile phone camera. The purpose of proposing this system is to automate attendance and p
2025-06-28 16:36:38 - Adil Khan
Vision Based Automated Attendance System Develop Using Deep Learning Techniques
Project Area of Specialization Artificial IntelligenceProject SummaryA-SYS is an android based mobile application for marking the attendance of students sitting in class. Attendance is marked by detecting the faces of students and the picture is captured through the instructor’s mobile phone camera. The purpose of proposing this system is to automate attendance and provide teachers with an efficient system to handle proxies and manual errors. This report contains detail information about all the systems. It defines that what this system does, how it performs its activities, how it stores data, relationship among the classes, brief and detail explanation of the of the events to which user will interact, flow of the system and the purpose of designing this system. This report has been created using the concepts of Ian Somerville and requirements are represented that is functional and non-functional using FURPS++ concept
Project ObjectivesTo launch an automated attendance system to address the traditional attendance system.
To remove proxy that exists in attendance systems.
To make a full-fledged record management system that maintains student’s attendance.
To send emails to students daily regarding attendance’s status and additional emails to those who are near or less to the threshold.
Project Implementation MethodThere are two modules of project web based and mobile based. Web is for admin where admin add the details of each student and faculty. Mobile is for faculty where faculty mark attendance of students. Both faculty and admin view records of student’s attendance weekly or monthly bases.
Attendance is marked by Face Recognition. Captured images send to server. There, first the faces of students extracted from captured image after that one by one each face will be recognized by his/her roll-number.
After that the faculty received a file of student who will be recognized then confirm and upload the final attendance.
After that emails send to all student on the bases of attendance’s status and also generate alert emails to those who are on edge of threshold.
Benefits of the Project- Use of ease for teachers to mark multiple student attendance at the same time.
- No false attendance
- Fast and accurate attendance marking system
Technical details for final deliverable are
- We use yolo V3 for Face Detection which give faster response and accuracy as compare to another algorithm.
- We use Convolution Neural Network for Face Recognition with maximum accuracy.
- We use volley request for APIs to get response from server side of attendance. All the user authentication is done using APIs.
- We use Html CSS for front end of Web and use JavaScript and C -sharp for back end
| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 122000 | |||
| Nvidia 1080 | Equipment | 1 | 20000 | 20000 |
| Smart Device with Best Camera Quality | Equipment | 1 | 30000 | 30000 |
| Plate Form and Device Compatibility Testing | Equipment | 1 | 5000 | 5000 |
| Deep Learning: Convolution Neural Networks In Python | Miscellaneous | 1 | 4000 | 4000 |
| Stationary | Miscellaneous | 1 | 2000 | 2000 |
| Nvidia 1080 | Equipment | 1 | 20000 | 20000 |
| Smart Device with Best Camera Quality | Equipment | 1 | 30000 | 30000 |
| Plate Form and Device Compatibility Testing | Equipment | 1 | 5000 | 5000 |
| Deep Learning: Convolution Neural Networks In Python | Miscellaneous | 1 | 4000 | 4000 |
| Stationary | Miscellaneous | 1 | 2000 | 2000 |