Fine-grained image recognition method and device, equipment and medium

The invention discloses a fine-grained image recognition method and device, equipment and a medium, and relates to the technical field of image recognition, and the method comprises the steps: training an initial attention twinning neural network constructed based on an attention mechanism and a twinning neural network through employing a pre-constructed training set, and obtaining a trained attention twinning neural network; pairing the to-be-classified pictures with any category of sample pictures in a training set, and inputting the paired pictures into an attention twinning neural network to obtain a prediction probability that the category of the to-be-classified pictures output by the attention twinning neural network and the category of the sample pictures belong to the same category; and determining a target category of the to-be-classified picture based on the prediction probability. According to the method, the attention mechanism and the twin neural network are combined to realize the category iden