Skip to main content
. 2019 Jul 1;26(11):1286–1296. doi: 10.1093/jamia/ocz089

Figure 3.

Figure 3.

Visualization of the proposed convolutional autoencoders based semisupervised learning model Convolutional AutoEncoder based Semi-supervised Network. The original images are first encoded into hidden codes through 5 layers of convolutional layers with a filter size of 4 × 4 and a stride of 2. The hidden codes can be either fed into fully connected layers for classification or into the 5 layers of deconvolutional layers for decoding into original images.