acquiring an image training sample data set; wherein the image training sample data set comprises image training sample data and tag information corresponding to the image training sample data;

constructing a basic activation function and a preset bias adjustment function as a preset activation function in an addition relationship, and determining the preset activation function to be an activation function of a neural network model, to obtain an initial neural network model; wherein, the preset bias adjustment function is a function constructed from a symbol function, a first trainable parameter and a quadratic term in a multiplication relationship;

inputting the image training sample data set into the initial neural network model for training until the model converges, to obtain a trained neural network model; and

when an image to be recognized is acquired, outputting a recognition result corresponding to the image to be recognized by using the trained neural network model,

wherein, the step of, constructing the basic activation function and the preset bias adjustment function as the preset activation function in the addition relationship, comprises:

constructing the basic activation function, the preset bias adjustment function and a preset linear function as an activation function in the addition relationship, to obtain the preset activation function;

wherein, the preset linear function comprises a second trainable parameter,

wherein, the step of, constructing the basic activation function, the preset bias adjustment function and the preset linear function as the activation function in the addition relationship, to obtain the preset activation function, comprises:

constructing the basic activation function, the preset bias adjustment function and the preset linear function as the activation function in the addition relationship according to a trainable weight parameter, to obtain the preset activation function,

wherein the preset activation function is:

wherein h(x) is the basic activation function, u(x) is the preset linear function, η(x) is the preset bias adjustment function, α is the trainable weight parameter, and

wherein, b and c are the second trainable parameters, and a is the first trainable parameter.