Sparsity-based image recognition model training method, device and equipment

The invention provides a sparsity-based image recognition model training method, device and equipment, and relates to the fields of computer vision, deep learning and the like in the technical field of artificial intelligence, and the method comprises the steps: inputting an obtained to-be-trained image into an initial training model; wherein the training model comprises a plurality of network layers used for processing image feature maps of the to-be-trained image, the network layers are configured with parameter matrixes, and the parameter matrixes are provided with sparseness; repeating the following steps until a preset stop condition is met: training the ith training model formed by the network layers with the jth-level sparseness to obtain the trained ith training model; according to the parameter matrix of the network layer of the ith training model after training, generating the network layer of the (i + 1) th training model; adding 1 to the value of i, and adding 1 to the value of j; and determining