Table 2. All layers and trainable parameters of the two-dimensional convolutional neural networks in this study.

Layer (type)	Output shape	Parameters #
Zeropadding2d_1	(None, 3, 22, 20)	0
Conv2d_1	(None, 1, 20, 32)	5,792
Max_pooling2d_1	(None, 1, 10, 16)	0
Zeropadding2d_2	(None, 3, 12, 16)	0
Conv2d_2	(None, 1, 10, 64)	9,280
Max_pooling2d_2	(None, 1, 5, 32)	0
Zeropadding2d_3	(None, 3, 7, 32)	0
Conv2d_3	(None, 1, 5, 128)	36,992
Max_pooling2d_3	(None, 1, 2, 64)	0
Zeropadding2d_4	(None, 3, 4, 64)	0
Conv2d_4	(None, 1, 2, 256)	147,712
Max_pooling2d_4	(None, 1, 1, 128)	0
Flatten_1	(None, 128)	0
Dense_1	(None, 256)	33,024
Dropout_1	(None, 256)	0
Dense_2	(None, 2)	514
Activation_1	(None, 2)	0