Sports Nutrition Intervention for Athletes Based on Continuous Image Deep Learning

Aiming at the problems of weak anti-noise ability, incompatibility of signal size and insufficient feature extraction in athlete sports nutrition intervention based on deep learning, a recognition method based on continuous image deep learning is proposed. Firstly, the time-varying signal is reconstructed into several continuous image frames to ensure that the input size is consistent; then a low-rank decomposition algorithm is designed to separate the key motion information annihilated by noise; at the same time, a depth model of time domain and spatial domain information fusion is proposed, Automatically capture the spatio-temporal characteristics of variable-length image sequences, and verify the proposed method on WiAR datasets and autonomously collected datasets. Experimental results show that the average recognition accuracy of the proposed method is 0.94 and 0.96, respectively, and has high accuracy and robustness in universal scenarios.