Two-dimensional convolution operation processing method, system and equipment and computer medium

Abstract

The invention discloses a two-dimensional convolution operation processing method, system and equipment and a computer medium. The method comprises the following steps of reading to-be-calculated dataand convolution kernel operation parameters from a preset memory; calling a first matrix conversion processor to convert the to-be-calculated data into target to-be-calculated data in a matrix form,and meanwhile calling a second matrix conversion processor to convert the convolution kernel operation parameters into target convolution kernel operation parameters in the matrix form; and determining a target convolutional layer in a preset convolutional layer group, and transmitting the target to-be-calculated data and the target convolutional kernel operation parameters to the target convolutional layer, so that the target convolutional layer performs two-dimensional convolutional operation on the target to-be-calculated data and the target convolutional kernel operation parameters, and the number of the same convolutional layers in the convolutional layer group is greater than or equal to 2. According to the two-dimensional convolution operation processing method provided by the invention, the efficiency of two-dimensional convolution operation can be improved. The two-dimensional convolution operation processing system and device and the computer readable storage medium also solve the corresponding technical problems.

Description

technical field [0001] The present application relates to the technical field of deep learning, and more specifically, relates to a two-dimensional convolution operation processing method, system, device, and computer storage medium. Background technique [0002] In deep learning technology, CNN (Convolutional Neural Networks, convolutional neural network) is mainly used as the research object, but due to different processing scenarios, the performance requirements of CNN are also different, which requires real-time adjustment of CNN according to the processing scenario, and The structure in CNN is mainly the convolution layer, so the efficiency of the two-dimensional convolution calculation of the convolution layer will affect the performance of CNN. [0003] To sum up, how to improve the efficiency of two-dimensional convolution calculation is an urgent problem to be solved by those skilled in the art. Contents of the invention [0004] The purpose of this application i...

AI Technical Summary

Problems solved by technology

[0004] The purpose of this application is to provide a two-dimensional convolution operation processing method, system, equipment and computer sto

Images