联合训练生成对抗网络的半监督分类方法

Deep neural networks require a large amount of data for supervised learning; however, it is difficult to obtain enough labeled data in practical applications. Semi-supervised learning can train deep neural networks with limited samples. Semi-supervised generative adversarial networks can yield superior classification performance; however, they are unstable during training in classical networks. To further improve the classification accuracy and solve the problem of training instability for networks, we propose a semi-supervised classification model called co-training generative adversarial networks (CT-GAN) for image classification. In the proposed model, co-training of two discriminators is applied to eliminate the distribution error of a single discriminator and unlabeled samples with higher confidence are selected to expand the training set, which can be utilized for semi-supervised classification and enhance the generalization of deep networks. Experimental results on the CIFAR-10 dataset and the SVHN dataset showed that the proposed method achieved better classification accuracies with different numbers of labeled data. The classification accuracy was 80.36% with 2000 labeled data on the CIFAR-10 dataset, whereas it improved by about 5% compared with the existing semi-supervised method with 10 labeled data. To a certain extent, the problem of GAN overfitting under a few sample conditions is solved.