Skip to main content
. 2021 Mar 11;2021:6695518. doi: 10.1155/2021/6695518

Table 2.

AlexNet 3D CNN model.

Layer Output shape Param (#)
Conv 3d_1 (conv 3D) (None 32, 32, 32, 16) 8208
Batch_normalization_v1 (None 32, 32, 32, 16) 64
Max_pooling 3D_1 (None 16, 16, 16, 16) 0
Conv3d_1 (Conv 3D) (None 16, 16, 16, 32) 13856
Batch_normalization_v1_2 (None 16, 16, 16, 32) 128
Max_pooling3d_2 (None 8, 8, 8, 312) 0
Conv3d_3 (None 8, 8, 8, 64) 55360
Batch_normalization_v1_3 (None 8, 8, 8, 64) 256
Conv3d_4 (None 8, 8, 8, 64) 110656
Batch_normalization_v1_4 (None 8, 8, 8, 64) 256
Conv3d_5 (conv 2D) (None 8, 8, 8, 32) 53328
Batch_normalization_v1_5 (None 8, 8, 8, 32) 128
Max_pooling2d_3 (None 4, 4, 4, 32) 0
Falatten_1 (flatten layer) (None, 2048) 0
Dense_1 (dense layer) (None, 200) 409800
Batch_normalization_v1_6 (None, 200) 800
Dropout_1 (dropout) (None, 200) 0
Dense_2 (dense layer) (None, 75) 15075
Batch_normalization_v1_7 (None, 75) 300
Dropout_2 (dropout) (None, 75) 0
Dense_3 (dense) (None, 2) 152