Research Article
Generation of Human Micro-Doppler Signature Based on Layer-Reduced Deep Convolutional Generative Adversarial Network
Table 2
Parameters of discriminator network.
| Row | Layer name | Layer type | Attribute |
| 1 | Input image | Image input | 64 × 64 × 3 images | 2 | Conv 1 | Convolutional | 64 conv filters of size 4 × 4 with stride [2, 2] and padding [1, 1] | 3 | Leaky ReLU 1 | Leaky ReLU | Scale of 0.2 | 4 | Conv 2 | Convolutional | 128 conv filters of size 4 × 4 with stride [2, 2] and padding [1, 1] | 5 | BN2 | Batch normalization | — | 6 | Leaky ReLU 2 | Leaky ReLU | Scale of 0.2 | 7 | Conv 3 | Convolutional | 256 conv filters of size 4 × 4 with stride [2, 2] and padding [1, 1] | 8 | BN3 | Batch normalization | — | 9 | Leaky ReLU 3 | Leaky ReLU | Scale of 0.2 | 10 | Conv 4 | Convolutional | 1 conv filter of size 8 × 8 with stride [1, 1] and padding [0, 0] |
|
|