Classroom student posture recognition based on an improved high-resolution network

被引:0
作者
Yiwen Zhang
Tao Zhu
Huansheng Ning
Zhenyu Liu
机构
[1] University of South China,School of Computer
[2] University of Science & Technology Beijing,School of Computer and Communication Engineering
[3] Hunan Provincial Base for Scientific and Technological Innovation Cooperation for Medical Big Data,undefined
来源
EURASIP Journal on Wireless Communications and Networking | / 2021卷
关键词
Pose estimation; Support vector machine; High-resolution networks; Squeeze-and-excitation networks; Object detection;
D O I
暂无
中图分类号
学科分类号
摘要
Due to the large number of students in a typical classroom and crowded seating, most features of student posture are often obscured, making it difficult to balance the accuracy in identifying student postures with computational efficiency. To solve this issue, a novel classroom student posture recognition method is proposed. First, to recognize the poses of multiple students in the classroom, we use the you-only-look-once (YOLOv3) algorithm for object detection and retrain it to detect human objects that are hunching on a table, creating the pose estimation network. Next, to improve the accuracy of the pose estimation network, we use the squeeze-and-excitation network structure that is embedded in the residual structure of high-resolution networks (HRNet). Finally, with the improved HRNet algorithm’s outputs of key human body points, we design a pose classification algorithm based on a support vector machine, to classify human poses in the classroom. Experiments show that the improved HRNet multi-person pose estimation algorithm yields the best mean average precision performance of 73.76% on the common objects in context (COCO) validation dataset. We further test the proposed algorithm on a customer dataset collected in a classroom and achieved a high recognition rate of 90.1% and good robustness.
引用
收藏
相关论文
共 12 条
[1]  
Tang L(2019)Pose detection in complex classroom environment based on improved Faster R-CNN IET Image Proc. 13 451-457
[2]  
Gao C(2017)Faster R-CNN: towards real-time object detection with region proposal networks IEEE Trans. Pattern Anal. Mach. Intell. 39 1137-1149
[3]  
Chen X(2013)Selective search for object recognition Int. J. Comput. Vis. 104 154-171
[4]  
Ren S(2019)Improved YOLOV3 object recognition algorithm with embedded SENet structure Comput. Eng. 45 243-248
[5]  
He K(undefined)undefined undefined undefined undefined-undefined
[6]  
Girshick R(undefined)undefined undefined undefined undefined-undefined
[7]  
Uijlings JRR(undefined)undefined undefined undefined undefined-undefined
[8]  
Sande KEAVD(undefined)undefined undefined undefined undefined-undefined
[9]  
Gevers T(undefined)undefined undefined undefined undefined-undefined
[10]  
Liu X(undefined)undefined undefined undefined undefined-undefined