Designing a Student Administration System Utilizing the YOLO-Based Model

The objective of this paper is to develop and implement a student administration system utilizing the YOLO model for information extraction from student cards, integrated into a web application interface using the FastAPI framework. Despite the relatively small size of our student card dataset and the presence of numerous small objects in close proximity within the student card images, we achieved a training mAP of over 83% and an inference of over 89% accuracy with 8-labeled classes. The positive testing results show how crucial it is to use object detection and recognition in a student administration system, providing the groundwork for implementing an automatic attendance system in schools later on.