摘要
针对舞蹈类在线课程教学过程中尚未实现教、学、评等有效闭环的问题,综合运用人工智能姿态识别算法MediaPipe和AphaPose,深入研究动姿智能评价与矫正技术,通过计算机深度学习训练,对舞蹈和运动动作的准确度进行评价,开发了该类课程的技术支持平台。平台对410个测评视频样本进行核验。结果表明,学员可用手机、平板电脑等智能终端采集动作练习视频,通过平台交互学习模块、智能评价与矫正模块,以图示或者动画演示等形式进行智能化的自动评价与矫正指导,可实现“导、学、练、评、纠”五大教学环节的深度融合,实现五维一体的健身教学服务。
In response to the issue of ineffective closed-loop teaching,learning,and evaluation in online dance courses,the artificial intelligence pose recognition algorithms MediaPipe and AphaPose were comprehensively applied to conduct indepth research on intelligent evaluation and correction techniques for posture.Through computer deep learning training,the accuracy of dance and motion movements was evaluated,and a technical support platform for this type of course was developed.The platform verified 410 evaluation video samples,and the results showed that students can collect action practice videos using smart terminals such as mobile phones and Pads.Through the platform's interactive learning module and intelligent evaluation and correction module,intelligent automatic evaluation and correction guidance can be carried out in the form of graphical or animated demonstrations.This can achieve deep integration of the five teaching links of"guidance,learning,practice,evaluation,and correction",and achieve a five dimensional integrated fitness teaching service.
作者
张尚先
李辉辉
王静
胡勇
ZHANG Shangxian;LI Huihui;WANG Jing;HU Yong(Guangdong Open University,Guangzhou Guangdong 510091,China;Zhongshan Polytechnic,Zhongshan Guangdong 528400,China;Guanghai Art Science and Technology Innovation(Shenzhen)Co.,Ltd.,Shenzhen Guangdong 518120,China)
出处
《信息与电脑》
2023年第15期244-247,共4页
Information & Computer
基金
广东省教育厅2021年继续教育质量提升工程项目“基于AI智能‘导、学、练、评、纠’一体的形体健身类在线课程研究与开发”(项目编号:继续教育教学改革与研究实践项目类第50号)
广东开放大学教育教学成果培育项目“基于AI线上线下融合的新体育课程开发与应用”(项目编号:2022LGCQ03-05)。
关键词
舞蹈在线课程
人工智能
技术平台
online dance courses
artificial intelligence
technology platform
作者简介
张尚先(1969-),男,湖南永州人,硕士研究生,教授。研究方向:教育管理;通信作者:李辉辉(1987-),男,广东梅州人,硕士研究生,讲师。研究方向:机电技术,智能控制。E-mail:804043038@qq.com。
