An Image Classification Method Based on Feature Recalibration Generative Adversarial Network

Abstract

An image classification method based on feature recalibration generative adversarial network, which is suitable for the field of machine learning. Input the image data to be classified into the confrontational network model for network training; the generator and the discriminator composed of the convolutional network; initialize the random noise, and input the random noise into the generator; use the convolutional network in the generator to perform multiple The layer deconvolution operation finally obtains the generated samples; the generated samples and the real samples are input into the discriminator; in the discriminator, the convolutional network is used to perform convolution and pooling operations on the input samples to obtain the feature map, and in the middle layer of the convolutional network The compressed activation SENet module is introduced to calibrate the feature map, the calibrated feature map is obtained, and the global average pooling is used to finally output the image data classification. The SENet module is introduced in the middle layer of the discriminator to automatically learn the importance of each feature channel, extract task-related and useful features, and suppress task-independent features, thereby improving the performance of semi-supervised learning.

Description

technical field [0001] The invention relates to an image classification method, and is especially suitable for an image classification method based on feature recalibration generation confrontation network. Background technique [0002] In the training process of a specific network, as the number of iterations increases, the network model often fits well in the training set and the training error is small, but the fitting degree of the test set is not good, resulting in a large test error. Current research shows that ensembles of multiple neural network models often perform better than a single neural network in the validation phase. In essence, it can be understood that the features extracted by different models for the same task are often different. This difference just makes up for the lack of generalization ability between models, so that the final task performance is better than a single model. many. [0003] The integration of discriminative models often combines mod...

Images