A classroom attention measurement scheme is proposed to address false and missed detections in head pose estimation within seminar or large-depth lecture classrooms. Firstly, a cascaded neural network is employed to detect heads in classroom videos in order to accurately locate and count student heads. Secondly, a global face detector and head pose classifier are developed to precisely determine gaze and non-gaze states by combining face detection and pose classification. Subsequently by introducing active parameters are introduced, and a head position matching algorithm is used to calculate head pose changes between frames. A new formula for head-up rate is derived based on gaze state changes. Finally, an attention model is analyzed to obtain an overall numerical value, the attention K-value, which combines head-up rate and active parameters. Experimental results indicate that the accuracy of the head detedor exceeds 95.6% in multi-scale classrooms, and the precision of the head classifier reaches 94.6%. Verification in real classrooms shows that the scheme accurately reflects student attention, providing valuable insights for teaching reflection and evaluation.
ZHANGY Y. Construction of a smart classroom for image processing courses in colleges and universities based on artificial intelligence: taking fundamentals of photoshop as an example[C]∥Proceedings of the 2023 IEEE 3rd International Conference on Social Sciences and Intelligence Management. Piscataway, USA: IEEE, 2023: 84-88.
[2]
XUX, TENGX. Classroom attention analysis based on multiple Euler angles constraint and head pose estimation[C]∥Proceedings of the International Conference on Multimedia Modeling. Cham, Switzerland: Springer, 2020: 329-340.
ZHENGZ, WANGP, LIUW, et al. Distance-IoU loss: faster and better learning for bounding box regression[C]∥Proceedings of the AAAI Conference on Artificial Intelligence. Washington, USA: AAAI, 2020:12993-13000.
[6]
LIW, WANGH Y, WANGR, et al. Reform and application of teaching design of experimental design course based on flipping classroom[C]∥Proceedings of the 2021 International Conference on Education,Information Management and Service Science. Piscataway, USA: IEEE, 2021:257-260.
[7]
TANGC K, LINH J, ZHANGL G, et al. Electrical specialty experiment teaching reform method with flipped classroom[C]∥Proceedings of the 2020 IEEE International Conference on Signal Processing, Communications and Computing. Piscataway,USA: IEEE. 2020:DOI:10.1109/ICSPCC50002.2020.925492 .
XIAOS H, SANGN, WANGX P, et al. Leveraging ordinal regression with soft labels for 3D head pose estimation from point sets[C]∥Proceedings of the 2020 IEEE International Conference on Acoustics, Speech and Signal Processing. Piscataway, USA: IEEE, 2020:1883-1887.
WANGQ L, WUB G, ZHUP F, et al. ECA-net: efficient channel attention for deep convolutional neural networks[C]∥Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway,USA: IEEE, 2020:11531-11539.
[12]
LIL, YAOD F. Emotion recognition in complex classroom scenes based on improved convolutional block attention module algorithm[J]. IEEE Access, 2023, 11:143050-143059.
[13]
FANGC X, LIUY F, DENGM. Research on team teaching model based on deep learning theory[C]∥Proceedings of the 2021 IEEE 3rd International Conference on Civil Aviation Safety and Information Technology. Piscataway,USA:IEEE, 2021:1234-1238.
[14]
XUEC H, ZHAOW, YUANT T, et al. Study on the classroom attention mechanism of deaf students based on three-in-one education model[C]∥Proceedings of the 2020 International Conference on Modern Education and Information Management. Piscataway,USA: IEEE, 2020:838-841.