Segmenting Brain Tissues from Chinese Visible Human Dataset by Deep-Learned Features with Stacked Autoencoder

Figure 3

Proposed three-hidden-layer SAE. Note that the number of layers in our model is set via cross-validation.