Small sample image classification method and system based on edge Transform graph neural network

Abstract

The invention discloses a small sample image classification method and system based on an edge Transform graph neural network, and the method comprises the steps: introducing a Transform model to update the edge features in a graph, dividing a difference feature graph between nodes into a feature block sequence, inputting the feature block sequence into a Transform module, and obtaining the updated edge features, and therefore, each pixel position can be allocated with different attention weights to highlight the key area. According to the idea of the method, a self-attention mechanism in transformer is used for automatically focusing to the key area for measuring the similarity between nodes, so that the purposes of inhibiting background information and highlighting the key area are achieved. A contrast experiment carried out on a miniImageNet data set proves that the method can improve the precision of small sample image classification.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to a small-sample image classification method and system based on an edge Transformer graph neural network. Background technique [0002] In recent years, thanks to the rapid improvement of computer computing power, deep learning has become a hot topic in the field of artificial intelligence research. However, behind the prosperity and development of deep learning is the support of large-scale manually labeled data sets, and the more complex the network, the more large data sets are needed for training. However, data in some special fields is very scarce, such as the discrimination of rare cases in medicine, and the existing limited medical images are far from enough to train a good deep model. At this time, it is hoped that the model can reduce its dependence on data and can learn quickly like humans, which will greatly reduce the cost of manual labeling of dat...

AI Technical Summary

Problems solved by technology

[0003]The existing small-sample learning model based on graph neural network aggregates node features based on global similarity. This method will aggregate a lot of background information and cause semantic In order to solve this problem, inspired by the transformer model, the present invention proposes a small-sample learning method that uses the transformer to perceive key areas, abandons the CNN and uses the transformer encoder to update the edge features in the GNN

Images