Creating a dataset to evaluate the condition of individuals in educational classrooms through in - depth study
Abstract
Assessing students' activity and participation levels during the educational process plays a crucial role in improving the quality of education. Using modern computer vision and artificial intelligence technologies, it is possible to automatically detect students' states in the classroom in real time. This research describes a system based on the LabelImg v1.8.1 annotation tool [6] and the YOLO (You Only Look Once) deep learning algorithm to identify student conditions such as hand-raising, sleeping, and phone usage. The proposed system is expected to successfully recognize student states with 90-95% accuracy, assisting teachers in enhancing lesson effectiveness
Keywords:
About the Authors
List of references
Nguyen, A., Chen, W., & Rauterberg, M. (2022). Detection of student attention levels in classroom environments using facial expressions analysis. IEEE Transactions on Learning Technologies, 15(3), 323–337.
Yang, L., Jiang, D., & Sahli, H. (2023). Posture analysis for student engagement assessment in classroom environments. International Journal of Human-Computer Studies, 167, 102709.
Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016). You only look once: Unified, real-time object detection. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 779–788).
Lin, T. Y., Goyal, P., Girshick, R., He, K., & Dollár, P. (2020). Focal loss for dense object detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, 42(2), 318–327.
Jocher, G., Stoken, A., & Borovec, J. (2023). YOLOv5 Documentation. Ultralytics. Retrieved from https://github.com/ultralytics/yolov5
Tzutalin. (2015). LabelImg. Git code. Retrieved from https://github.com/tzutalin/labelImg
Yang, F., & Wang, T. (2024). SCB-Dataset3: A Benchmark for Detecting Student Classroom Behaviour. arXiv:2310.02522v2 [cs.CV]. 9 Sep 2024.