基于Y0L0v11与人脸识别的课堂举手行为智能检测系统研究

打开文本图片集

中图分类号：TP311 文献标识码：A 文章编号：2096-4706（2025）24-0033-06

Abstract：Withthecontinuousadvancementandin-depthappicationofArtificial Intelligence，clasroombehaviordetection has become animportant directionin thefieldofsmart education.Toaddressissuessuch aslowaccuracyinclassroombehavior detectioandlack of identityannotation，thispaper designs andimplements ahand-raising behaviordetectionsystem integrating YOLOvllandaceogitiooporatigbhaiordetectio，entitycogitioadvisualatiomoulestupporta recognition and individual binding. On SCB-Dataset5，the optimized model achieves a mAP（∅0.5 of 82.7% and a m AP@0.5：0.95 （20 of 64.4% .In actual classroom scenarios，the hand-raising detection accuracy reaches 68.3% ，and the face recognition accuracy is 65.0% .This research provides technical support for classroom behavior analysis and teaching applications.

Keywords： YOLOvl1; face recognition; classroom behavior detection; Deep Learning; smart education

0 引言

随着人工智能与计算机视觉技术的持续发展，智慧教育正加速从理念迈向实践[1。（剩余7585字）