Expression recognition system and method based on two-stage self-healing network

The invention provides an expression recognition system and method based on a two-stage self-healing network, and relates to the technical field of image recognition processing. Acquiring a face image and processing the face image into a standard format; face features are extracted from the to-be-recognized face image through a feature extraction algorithm, the face features serve as input, gender types corresponding to the to-be-recognized face image are recognized through the expression recognition model, and parameters in the expression recognition model are frozen after training is completed; based on the frozen parameter value, identifying an expression type corresponding to the to-be-identified face image through an expression identification model; wherein the expression recognition model is obtained by performing deep learning training through a large amount of expression data sets and adjusting expression recognition model parameters in the training process; according to the expression recognition mod