Table 1 MFR-Net V1 network architecture
From: Coverless image steganography using morphed face recognition based on convolutional neural network
Type | Output size |
|---|---|
Input | 224 × 224 |
Con2D 7 × 7, 64, stride 2 Batch normalization leaky ReLU | 112 × 112 |
Max pool, 3 × 3, stride 2 | 56 × 56 |
(Residual block, 64) × 3 |  |
(Residual block, 128) × 4 | 28 × 28 |
(Residual block, 256) × 6 | 14 × 14 |
(Residual block, 512) × 3 | 7 × 7 |
Global average pool | 512 |
FC L2 norm, 300-d | 300 |
Softmax | Â |