Digital Audio Scene Recognition Method Based on Machine Learning Technology

Table 1

Parameters of convolutional neural network.

Layer (type)	Output size	Parameter
Conv_1	(40, 500, 128)	6400
Batch standardization_1	(40, 500, 128)	160
Activation function_1	(40, 500, 128)	0
Maximum pooling_1	(8, 100, 128)	0
Dropout_1	(8, 100, 128)	0
Conv_2	(8, 100, 256)	1605888
Batch standardization_2	(8, 100, 256)	32
Activation function_ 1	(8, 100, 256)	0
Maximum pooling_1	(2, 1, 256)	0
Dropout_1	(2, 1, 256)	0
Dense (full connection layer)_1	100	51300
Dropout_3		0
Dense (full connection layer)_2	100	1515