Table 1 Description of network architectures used for experimentation
From: Face recognition with Bayesian convolutional networks for robust surveillance systems
Layer No. | Architecture # 1 (Arc-1) | Architecture # 2 (Arc-2) | Architecture # 3 (Arc-3) | |||
|---|---|---|---|---|---|---|
Layer type @ size | Activation | Layer type @ size | Activation | Layer type @ size | Activation | |
1 | C1 @ 3 × 3 × 32 | ReLU + dropout | C1 @ 3 × 3 × 32 | ReLU + dropout | C1 @ 3 × 3 × 32 | ReLU + dropout |
2 | P1 @ 2 × 2 | – | P1 @ 2 × 2 | – | P1 @ 2 × 2 | – |
3 | C2 @ 3 × 3 × 32 | ReLU | C2 @ 3 × 3 × 32 | ReLU | C2 @ 3 × 3 × 32 | ReLU |
4 | P2 @ 2 × 2 | – | P2 @ 2 × 2 | – | P2 @ 2 × 2 | – |
5 | C3 @ 3 × 3 × 64 | ReLU | C3 @ 3 × 3 × 64 | ReLU | C3 @ 3 × 3 × 32 | ReLU |
6 | P3 @ 2 × 2 | – | P3 @ 2 × 2 | – | P3 @ 2 × 2 | – |
7 | FC1 @ 512 | ReLU + dropout | C4 @ 3 × 3 × 64 | ReLU | C4 @ 3 × 3 × 64 | ReLU |
8 | FC2 @ 128 | ReLU + dropout | P4 @ 2 × 2 | – | P4 @ 2 × 2 | – |
9 | FC3 | Softmax | FC1 @ 512 | ReLU + dropout | C5 @ 3 × 3 × 64 | ReLU |
10 | – | – | FC2 @ 128 | ReLU + dropout | P5 @ 2 × 2 | – |
11 | – | – | FC3 | Softmax | FC1 @ 512 | ReLU + dropout |
12 | – | – | – | – | FC2 @ 128 | ReLU + dropout |
13 | – | – | – | – | FC3 | Softmax |