A semi-supervised learning model based on convolutional autoencoder and convolutional neural network for image classification
Yuxuan Li, Hsiang‐Yuan Yeh · 2019
Deep learning has achieved the state-of-the-art performance in image classification. But, the model with supervised learning approach should be trained with large parameters and completely labeled datasets. Therefore, we proposed a semi-supervised learning model based on a convolutional auto-encoder and a complementary convolutional neural network to assist image classification. Experimental results show that in the proposed model, the number of labelled data can be reduced by more than half, and the classification accuracy continues to have the same performance. The results show that the effectiveness and feasibility of our model with a limited number of labeled data.