Personalized federal learning method, device and system based on global feature sharing

The invention relates to a personalized federal learning method, device and system based on global feature sharing. The personalized federal learning method based on global feature sharing is applied to a client, and comprises the following steps: receiving a global feature extractor model and global features sent by a server; initializing a local model according to the global feature extractor model and the local classifier model; the local image data are input into the initialized local model for model training, a loss function of the local model is determined, and the loss function comprises cross entropy loss between a training label and a real label of the local image data and a conditional mutual information regular term; according to the loss function of the local model, performing first updating processing on the local model based on back propagation; and when the local model converges, determining a target local model. According to the method, the global features and the conditional mutual informatio