Artificial intelligence technology as an important force driving the new round of scientific and technological revolution and industrial change, Open Pose Skeletal Tracking is a kind of algorithm based on computer vision technology, which plays an important role in quantitative sports and whole-body posture detection. In this paper, we constructed an intelligent system based on Open Pose bone tracking technology to assist aerobics training, and determined the corresponding experimental indexes, and carried out an 8-week experimental intervention on 50 athletes through the experimental method. The study concludes that artificial intelligence can be an empowering aid for performance-oriented and differentiated training-based athletic training, which is currently faced with the obstacles of insufficient infrastructure, privacy leakage of athletes, and incomplete talent protection mechanism.