Research on the Enhancement Path of Physical Education Teaching in Colleges and Universities under Artificial Intelligence Technology

Abstract Smart sports teaching is regarded as a new type of education mode, aiming to improve the quality and effect of sports teaching through the use of artificial intelligence technology. The article first combined the opportunities for digital transformation of college sports teaching given by AI technology, combined the application scenarios of AI technology to innovative college sports teaching, and established a smart sports teaching platform with smart IoT architecture. The OpenPose algorithm is utilized to extract the key points of the skeletal movements of sports teaching, and the optimal matching planning is carried out by the DTW algorithm combined with the ICP algorithm for the movement postures of sports. For the effectiveness of the algorithm, corresponding identification verification experiments were designed, and teaching comparison experiments were created on the intelligent sports teaching platform. The OpenPose algorithm spent only 0.284s in acquiring the skeletal key points of sports, and the average success rate of matching the movement postures of sports was 89.75%. There was a maximum difference of 10.81 points in the stage movement assessment between students of class A and class B, and the negative interest of students of class A in physical education was reduced by 40.97%. Negative interest decreased by 40.97%. Relying on the intelligent sports teaching platform and combining artificial intelligence technology to technologically upgrade sports teaching in colleges and universities provides a new path to enhance the quality of sports teaching in colleges and universities.