Digital Signal Processing and Artificial Intelligence for Automatic Learning Jianli Wang, Ruichun Gu 3 4 2024 https://doi.org/10.6025/dspaial/2024/3/4/121-127 https://www.dline.info/dspai/fulltext/v3n4/dspaiv3n4_1.pdf This paper mainly studies the application of computer neural network algorithms in the training and testing of upper and lower limb coordination in aerobics. By analyzing deep learning algorithms, we can accurately assess and predict the coordination ability of aerobics athletes, thereby improving their competitiveness. We introduce the application of computer neural network algorithms in sports training and testing, especially for testing upper and lower limb coordination in aerobics. We use deep learning algorithms, such as convolutional neural networks (CNN), to evaluate and predict the coordination of athletes’ upper and lower limbs. Our research results show that computer neural network algorithms have broad application prospects in the training and testing upper and lower limb coordination in aerobics.