Table 1 Parameters in this paper.
From: Brain CT image classification based on mask RCNN and attention mechanism
Layer | Parameter | Output size |
|---|---|---|
Conv1 | 7 × 7, 64, stride = 2 | 100 × 100 |
Max Pool | 3 × 3, stride = 2 | 50 × 50 |
Residual Unit 1 | 3 × 3, 64 | 50 × 50 |
Attention model | RHAM | 50 × 50 |
Residual Unit 2 | 3 × 3, 128 | 25 × 25 |
Attention model | RHAM | 25 × 25 |
Residual Unit 3 | 3 × 3, 256 | 13 × 13 |
Attention model | RHAM | 13 × 13 |
Residual Unit 4 | 3 × 3, 512 | 7 × 7 |
Attention model | RHAM | 7 × 7 |
Global average pool | 1 × 1 | |
Dropout | 0.5 | |
FC, softmax | 3 | |
