Multi-label image classification method and system based on class activation mapping mechanism

Abstract

The invention discloses a multi-label image classification method and system based on a class activation mapping mechanism, and the method comprises the steps: obtaining a to-be-classified image, and converting the to-be-classified image into a multi-dimensional tensor; inputting the multi-dimensional tensor into a classification model to obtain a classification result of the to-be-classified image; in the model training stage, the label of each image is firstly converted into a label word vector, then the co-occurrence relationship between different labels is learned, and the relationship is fused into the label word vector, so that the problem that the dependency relationship between the labels cannot be fully learned by the existing image classification method, and the technical problem of poor image classification effect is solved. Besides, due to the fact that a class activation mapping mechanism is combined in the model training stage, the visual consistency of different styles of the same image is maintained, the classification effect of the model is improved, and the method can be widely applied to the technical field of artificial intelligence.

Description

technical field [0001] The invention relates to the technical field of artificial intelligence, in particular to a multi-label image classification method and system based on a class activation mapping mechanism. Background technique [0002] Nowadays, Multi-label image classification (Multi-label image classification) has been widely used in the field of computer vision, including multi-target recognition, sentiment analysis, medical diagnosis recognition, etc. Since each image contains multiple objects, and an image contains multiple styles, how to effectively learn the relationship between these objects and how to maintain the visual consistency between different styles of the same image is still full of challenges. challenging. [0003] Generally speaking, the existing multi-label image classification methods first use the convolutional neural network to obtain the feature vector of the image, then use the graph convolutional network (Graph Convolutional Network, GCN) t...

AI Technical Summary

Problems solved by technology

[0003] Generally speaking, the existing multi-label image classification methods first use the convolutional neural network to obtain the feature vector of the image, then use the graph convolutional network (Graph Convolutional Network, GCN) to obtain the co-occurrence relation word vector between the labels, and finally directly The dot product operation of the vector is used to fuse the image features and the co-occurrence relationship word vector of the label, but it does not consider the visual consistency between different styles of the image, which affects the image recognition performance of the image classification model.

Images