Table 2.

Network hyperparameters.

Training Parameters	Values
Image Size	640*640
Epochs	250
Batch Size	64
Learning Rate	0.01
Optimizer	SGD
Momentum	0.937
Weight Decay	0.0005