Abstract: A method for optimizing operation of convolution neural network outputs an input matrix of an input image. The method further slides on the input matrix according to preset convolution kernels to perform dot products to output first output matrixes on a convolution computing layer. Nonlinear mapping of the first output matrixes is performed according to a preset activation functions on an activation layer to output a second output matrixes. The method does not perform bias operations on the convolution computing layer. The results of the dot products performed on the convolution computing layer are output to the activation layer and bias operations are performed on the activation layer. The method reduces amount of calculation whilst not reducing accuracy of processing. An electronic device, a convolution neural network, and a non-transitory storage medium are also disclosed.

Abstract: An image recognition method using reduced processes of calculation acquires training data and trains parameters of fully connected layers of a neural network model with the data. The neural network model is an image recognition model, the training of the parameters of fully connected searches for nodes of at least one connected layer and divides the nodes in a same layer of the into multiple groups according to a first preset rule. A feature node in each of the multiple groups is determined according to a second preset rule and feature node information in each group is used as parameter information to complete the training of the fully connected layer. An image is acquired and input in response to receiving an image recognition command, and an image recognition result is output. A computing device applying the method is also disclosed.