Table 3 Discriminator architecture: Minibatch Size: 32; Optimizer: Adam (η = 0.0001, β1 = 0.5, β2 = 0.999).
Layer1 | Layer2 | Layer3 | Layer4 | Layer5 | Output | |
|---|---|---|---|---|---|---|
Type | Convolution | Convolution | Convolution | Convolution | Convolution | Full Con |
Input | [32 × 32 × 2] | [64 × 64 × 64] | [32 × 32 × 128] | [16 × 16 × 256] | [8 × 8 × 512] | [4 × 4 × 1024] |
Output | [64 × 64 × 64] | [32 × 32 × 128] | [16 × 16 × 256] | [8 × 8 × 512] | [4 × 4 × 1024] | [1] |
Activation | LeakyReLU | LeakyReLU | LeakyReLU | LeakyReLU | LeakyReLU | Sigmoid |
Batch norm | Yes | Yes | Yes | Yes | Yes | – |
Stride | 2 | 2 | 2 | 1 | 1 | – |
Padding | Same | Same | Same | Same | Same | – |
Kernel Size | 5 | 5 | 5 | 5 | 5 | – |
Kernels | 64 | 128 | 256 | 512 | 1024 | – |
