Image classification method and device

Abstract

One or more embodiments of the invention provide an image classification method and device, and the method comprises the steps: building a residual network model, replacing a standard convolution in an original side diameter of the residual network model with a cavity convolution, and generating a cavity residual network trunk; generating a weight layer of the residual error network model based ona channel attention module and a space attention module of an attention mechanism model; generating a residual attention mechanism model composed of the cavity residual network trunk and the weight layer, and training the residual attention mechanism model; and inputting image data into the residual attention mechanism model, and identifying and classifying the image data. According to one or more embodiments of the present specification, a residual mechanism is integrated into the attention model, and contextual information inside the attention mechanism is combined without increasing parameters, so that the attention model is helped to extract features of interest of an image classification task more accurately, and the efficiency and accuracy of image classification are improved.

Description

technical field [0001] One or more embodiments of this specification relate to the technical field of image recognition, and in particular, to an image classification method and device. Background technique [0002] With the improvement of social informatization, images have gradually replaced text as an important carrier for human transmission and storage of information. The disorder and huge volume of information contained in images pose a huge challenge to the processing of image information. How to effectively classify images to extract the useful information we need has become a topic of much concern in the field of computer vision. [0003] However, with the development of society, the amount of image data is increasing exponentially, and the scope of application is constantly expanding. The network structure and algorithm of image classification in the existing technology are far from perfect and efficient for different types, different natures and disorderly images....

AI Technical Summary

Problems solved by technology

[0004] In view of this, the purpose of one or more embodiments of this specification is to propose an image classification method and equipment to solve the problem of low image classification efficiency and accuracy

Images