Development of Deep Learning Model for the Recognition of Cracks on Concrete Surfaces

Table 2

Parameters during the training phase.

Parameter	Value and description
Initial learning rate	0.01
Learning rate schedule	Constant
Momentum	0.9
Gradient threshold method	l2 norm
Maximum number of epochs	10
Iteration per epoch	156
Maximum number of iterations	1560
Validation frequency	1 iteration
Shuffle the data	Every epoch