Skip to main content
. 2021 Mar 30;11(4):616. doi: 10.3390/diagnostics11040616

Figure 3.

Figure 3

The architecture of the VGG-16- and VGG-19-CXR-U-Net models. The only difference is that the VGG-19-CXR-U-Net has an additional Conv-512-BN-ReLU layer in the convolutional blocks 3, 4, and 5 in the encoder path.