Table 5 Comparison of computational cost.
From: Underwater vessel sound recognition based on multi-layer feature and attention mechanism
Method | Modules | Flops (M) | Accuracy | GPU (MB) | Parameters (M) | Time (s) |
|---|---|---|---|---|---|---|
EDAFF-TDNN | Without feature fusion, SE block, dilated convolution | 42.86 | 96.3 | 1558 | 5.35 | 26 |
With feature fusion, SE block, dilated convolution | 49.17 | 98.2 | 1802 | 6.14 | 28 |
