Lightweight classification method of human action in long video of basketball competition

Aiming at the problem that the current lightweight classification effect of human motion is poor, this paper puts forward the lightweight classification method of human motion in the long video of basketball sports competition, collects and recognizes the characteristics of human motion in the long video of basketball sports competition combined with the principle of neural network, and further uses the deep learning model to lightweight classify the motion characteristics. Finally, it is confirmed by experiments, The lightweight classification method of human action in long video of basketball sports competition has high practicability and accuracy. It can better know that basketball players carry out sports training and fully meet the research requirements.