| Idea Name: |
Smart Attendance |
| Slogan: |
Lets |
| Supervisor Name: |
Dr. Usman Zabit |
| Supervisor Designation: |
Associate Professor |
| Supervisor School: |
SEECS |
| Supervisor Department: |
Department of Electrical Engineering |
| Contact number: |
051-9085-2122 |
| Email ID: |
[email protected] |
| Abstract: |
| Face Recognition technologies have made many improvements in the world. Smart Attendance using Real-Time Face Recognition is a real-world solution to the traditional handling of student attendance. |
| What is the unmet need in society that your idea will fulfill ? |
| One of the major challenges in a smart classroom is to develop an unobtrusive attendance management system. Traditional attendance marking systems interrupt the teaching-learning process, consume a lot of time, and are prone to student proxies. Smart Attendance offers a streamlined way to do this. |
| Who needs it ? How many would benefit ? |
| Smart Attendance finds it application in schools & universities. Slight modifications to the system can enable its use in offices, surveillance for security as well as event check-ins currently relying on passes/tickets etc. |
| How will the solution works |
| The project employs state of the art face recognition techniques based on Deep Neural Networks. An image dataset of students will be used to ‘train’ the model to recognize student faces. Digital video streaming cameras installed in classrooms will capture entire-class images in one frame, which will be passed though the trained model. The recognized students will be marked present and attendance will be communicated to a database maintaining the record. |
| Who are your competitors ? How is your solution different |
| Our competitors utilize obtrusive recognition methods like RFID based cards, bio-metric systems like IRIS and fingerprint recognition. In contrast to existing face recognition systems which perform time consuming singular recognitions, our system is unobtrusive & eliminates any class interruptions. |
| Status: |
new |
| Entry Date & Time: |
2020-12-20 (1140) |
