US Patent 11829476 Computing device and model parameters security protection method

A model parameters security protection method is implemented in a computing device in communication connection with at least one security protection device. The method includes training a data model based on an artificial neural network using a number of images and obtaining parameter information of the data model, encrypting the parameter information and generating a configuration file comprising the encrypted parameter information, and sending the configuration file to the at least one security protection device. The parameter information includes at least one of a weight of neuron and an offset value of the neuron of the artificial neural network.