A

Automated Face Recognition and Attendance Project

Automated Face Recognition and Attendance project is a prototype of an attendance-monitoring system that uses facial recognition that is a practical replacement of other general methods

Created on 9th October 2021

A

Automated Face Recognition and Attendance Project

Automated Face Recognition and Attendance project is a prototype of an attendance-monitoring system that uses facial recognition that is a practical replacement of other general methods

The problem Automated Face Recognition and Attendance Project solves

Doesn't matter if it is a class full of students , or an office full of employees , every time someone has to take care of attendance of others. This is not only a lengthy(time-consuming) process but also interrupts the regular flow. At the same time attendance sheet can get damaged ( is some cases loss) during the process. And when the number od attendees grows higher it might end up in chaos.
Not in pen-paper attendance, even in case of bio-metric attendance employees have to wait making a queue during rush hours to get their chance. Magnetic Chip-based attendance system faces the same issue.
But all of these problems can be solved if the employees doesn't have to wait in queue or neither another person needs to be recruited to take attendance if we use facial-recognition based attendance system.
The attendance system presented by team Gigabyte solves the problems stated above. A camera installed in right place can capture the face of employee and takes log of attendance.

Challenges we ran into

While solving the problem statement we ran into some challenges but our team-effort made us sail through all of them. A few of them are -
1. It was out first project on facial-recognition and Deep Learning. So before we could come up with any solution we had to check through a lot of research-papers, blogs, projects.
2. Our project consists of many python libraries ( such as - dlib, cv2, numpy,face_recognition). Before we could code anything we had to learn all of them first and mane different functions available in those libraries.
3. When it comes to face-recognition it is not easy to find proper deep-learning algorithms as it has to be trained through thousands of pictures to make it efficient and also less time consuming. Thankfully face_recognition library of python comes handy in this case. It uses HOG algorithm to find out specific 128 encodings for each and every face.

Discussion

Builders also viewed

See more projects on Devfolio