Employing Artificial Intelligence Technologies To Assess Motor Performance Of High Jump Among Students

Abstract

The high jump is one of the most biomechanically complex athletics events, requiring a precise integration of approach speed, takeoff power, and motor coordination during the jump phase. While assessing motor performance is crucial for improving achievement, traditional evaluation methods often lack objectivity. Sufficient accuracy. This research aims to employ modern artificial intelligence techniques (OpenPose and BlazePose) to objectively and accurately assess students' motor performance in the high jump. The researcher used the descriptive analytical method on a sample of 35 students in Faculty of Physical Education and Sports Sciences The fourth phase was conducted at the universities of Wasit, Kut, and Qadisiyah . Performance phases (approach, take-off, and crossbar) were filmed with high-speed cameras (240 frames per second), and computer vision techniques were used to automatically extract kinematic variables. A deep neural network model with an 18-12-8-1 architecture was built to evaluate performance based on 18 biomechanical variables, with the results compared to the evaluations of three expert specialists. The results showed that the intelligent model achieved a very strong correlation coefficient (r=0.93, p<0.001) with expert evaluations, a high reliability coefficient (r=0.91) upon test retesting, and an excellent objectivity coefficient (r=0.94). The analysis also revealed that approach speed (r=0.87), take-off angle (r=0.82), and height of center of gravity upon clearing the bar (r=0.89) were the three most influential variables. Five graded standard levels were developed specifically for students in faculties of physical education and sports science (weak, acceptable, good, very good, and excellent). The research recommended adopting this technology in the practical courses of athletics in physical education colleges, training students and faculty members on its use, and developing a smartphone application that facilitates the use of the system in the educational environment.