Table 1 Experimental parameters.
From: A target detection model HR-YOLO for advanced driver assistance systems in foggy conditions
Parameter name | Parameter value |
|---|---|
Initial Learning rate | 0.01 |
Batchsize | 16 |
Momentum | 0.937 |
Training Epoch | 300 |
Decay strategy | Cosine annealing decay |
Final learning rate | 0.0001 |
Data augmentation | Mosaic data augmentation |
Closing Mosaic epochs | 10 |
Weight decay | 0.0005 |
