The Application of Athlete Potential Evaluation System in Student Physical Training
In response to the problem that the current athlete evaluation system cannot explore students'potential physical fitness,this article combines video analysis technology,LSTM technology,Gaussian mixture,visual spatiotemporal evaluation and other methods to develop a system suitable for athlete ability prediction and evaluation.The system extracts meta events from first person videos for predictive analysis and constructs a highly nonlinear visual space.Through convolutional neural network learning,it evaluates the behavioral characteristics of athletes.The experimental results show that the evaluation method proposed in this paper has higher prediction accuracy compared to other first person perspective evaluation models.It can effectively identify the contribution behavior and negative behavior of students'physical training,and use this as a basis to predict and evaluate students'physical potential.
evaluate performanceprediction systemLSTM technologyconvolutional networksstudent physical training