Table 2 The network parameters of the backbone network
From: AGV monocular vision localization algorithm based on Gaussian saliency heuristic
Level | Size/filter | Pool size | Stride | Padding | Output (input) |
|---|---|---|---|---|---|
Input layer | – | 1 | |||
Layer 1 | 3 × 3/6 | 2 | 1 | 1 | 6 |
Layer 2 | 3 × 3/12 | 2 | 1 | 1 | 12 |
Layer 3 | 3 × 3/20 | 2 | 1 | 1 | 20 |
Layer 4 | 3 × 3/40 | 2 | 1 | 1 | 40 |
Layer 5 | 3 × 3/64 | 2 | 1 | 1 | 64 |
Layer 6 | 3 × 3/128 | 2 | 1 | 1 | 128 |
Layer 7 | 3 × 3/256 | 2 | 1 | 1 | 256 |