Research Article

Local Binary Convolutional Neural Networks' Long Short-Term Memory Model for Human Embryos' Anomaly Detection

Table 7

Baseline CNN architecture details.

Layer (type)Output shapeConnected to

Input (input layer)(1024, 1024, 3) 
conv2d_1 (Conv2D)(1024, 1024, 3)Input
conv2d_2 (Conv2D)(1024, 1024, 3)conv2d_1
conv2d_3 (Conv2D)(512, 512, 512)conv2d_2
conv2d_4 (Conv2D)(512, 512, 512)conv2d_3
conv2d_5 (Conv2D)(512, 512, 512)conv2d_4
conv2d_6 (Conv2D)(512, 512, 512)conv2d_5
conv2d_7 (Conv2D)(512, 512, 512)conv2d_6
add_1 (Add)(512, 512, 512)conv2d_4 and conv2d_7
conv2d_8 (Conv2D)(256, 256, 512)add_1
conv2d_9 (Conv2D)(256, 256, 256)conv2d_8
conv2d_10 (Conv2D)(256, 256, 256)conv2d_9
conv2d_11 (Conv2D)(256, 256, 512)conv2d_10
conv2d_12 (Conv2D)(256, 256, 256)conv2d_11
add_2 (Add)(256, 256, 256)conv2d_8 and conv2d_12
conv2d_13 (Conv2D)(128, 128, 512)add_2
conv2d_14 (Conv2D)(128, 128, 128)conv2d_13
conv2d_15 (Conv2D)(128, 128, 128)conv2d_14
conv2d_16 (Conv2D)(128, 128, 512)conv2d_15
conv2d_17 (Conv2D)(128, 128, 128)conv2d_16
add_3 (Add)(128, 128, 128)conv2d_13 and conv2d_17
conv2d_18 (Conv2D)(64, 64, 512)add_3
conv2d_19 (Conv2D)(64, 64, 64)conv2d_18
conv2d_20 (Conv2D)(64, 64, 64)conv2d_19
conv2d_21 (Conv2D)(64, 64, 512)conv2d_20
conv2d_22 (Conv2D)(64, 64, 64)conv2d_21
add_4 (Add)(64, 64, 64)conv2d_18 and conv2d_22
Average pool (GlobalAveragePooling2)(64)add_4