Skip to main content

Table 1 Details of the discriminator network

From: Superresolution reconstruction method for ancient murals based on the stable enhanced generative adversarial network

Name Type Kernel Stride Padding Output
Conv0_0 conv 3 × 3 1 × 1 1 64
Conv1_0 conv 4 × 4 2 × 2 1 64
Conv1_1 conv 3 × 3 1 × 1 1 64
Conv2_0 conv 4 × 4 2 × 2 1 128
Conv2_1 conv 3 × 3 1 × 1 1 128
Conv3_0 conv 4 × 4 2 × 2 1 256
Conv3_1 conv 3 × 3 1 × 1 1 256
Conv4_0 conv 4 × 4 2 × 2 1 512
Conv4_1 conv 3 × 3 1 × 1 1 512
FC0 FC 100
FC1 FC 1