Skip to main content

Table 1 The network architecture

From: Automated cardiovascular magnetic resonance image analysis with fully convolutional networks

Scale Size Convolution
1 192×192 3×3, 16
   3×3, 16
2 96×96 3×3, 32
   3×3,32
3 48×48 3×3, 64
   3×3, 64
   3×3, 64
4 24×24 3×3, 128
   3×3, 128
   3×3, 128
5 12×12 3×3, 256
   3×3, 256
   3×3, 256
upsample and concatenate scale 1 to 5 features  
predict 192×192 1×1, 64
   1×1, 64
   1×1, K
  1. The first two columns list the resolution scale and feature map size. The third column lists the convolutional layer parameters, with “ 3×3,16” denoting 3×3 kernel and 16 output features. The last convolutional layer outputs K features, with K denoting the number of label classes