Reforming university physical education with artificial intelligence for enhanced teaching and learning

Abstract Being in a rapidly changing educational environment, the emergent technology of AI seems to have played the disruptive role in converting the conventional methodology of teaching in each subject. PE is being challenged through the need for the conventional systems to monitor physical parameters of children in real time-the one-needs-for-individual-feedback-and-performance-analysis. A transforming factor for conventional teaching being renewed by AI with respect to other subjects in the rapidly evolacing on educational landscape, PE finds difficulty with the conventional systems being there to monitor children's physical states in real time, preventing these systems from affording individual feedback and performance breakdowns on the children’s implement this recognition system, a dataset was created using wearable sensor technology and motion capture based on video footage. The collected data underwent preprocessing, which involves removing errors by filtering out noise using signal processing techniques. The Short-Time Fourier Transform is applied so that the local frequency information in time can be acquired, with the objective of improving motion feature representation. The experimental findings indicate that the proposed model yields accuracy (98.6%), precision (97.9%), recall (96.3%), and F1 score (97.9%) that are better than those of traditional algorithms, with high recognition precision in motion analysis. The findings indicate that university PE driven by AI generates an interactive, data-centric, and adaptive learning platform, significantly enhancing students' performance and teachers' efficiency. The room for becoming of AI is seen in having a voice for furthering the possibilities of instruction in another dimension on physical education at the college level, where there is personalization and technology-based instruction.