Computational Models-Based Detection of Peripheral Malarial Parasites in Blood Smears

Table 3

Pretrained convolution neural network model (VGG-19).

Layer	Output	Parameter
input_2	[(N, 125, 125,3)]	0
b1-c1	(N, 125, 125, t1)	P1
b1-c2	(N, 125, 125, t1)	P1
b1-pool	(N, 62, 62, t1)	0
b2-c1	(N, 62, 62, t2)	P2
b2-c2	(N, 62, 62, t2)	P2
b2-pool	(N, 31, 31, t3)	0
b3-c1	(N, 31, 31, t4)	P3
b3-c2	(N, 31, 31, t4)	P3
b3-c3	(N, 31, 31, t4)	P3
b3_c4	(N, 31, 31, t4)	P3
b3-pool	(N, 15, 15, t4)	0
b4-c1	(N, 15, 15, t5)	P4
b4-c2	(N, 15, 15, t5)	P4
b4-c3	(N, 15, 15, t5)	P4
b4-c4	(N, 15, 15, t5)	P4
b4-pool	(N, 7, 7, t5)	0
b5-c1	(N, 7, 7, t5)	P5
b5-c2	(N, 7, 7, t5)	P5
b5-c3	(N, 7, 7, t5)	P5
b5-c4	(N, 7, 7, t5)	P5
b5-pool	(N, 3, 3, t5)	0
fla_1	(N, t5)	0
Dse-3	(N, dset5)	P5
Dt-2	(N, dttt5)	0
Dse-4	(N, dset5)	262656
Dt-3	(N, dt5)	0
Dse-5	(N, 1)	513
Total params:		22,647,361
Trainable params:		2,622,977
Nontrainable:		20,024,384

Note: b1-c1 and b1-c2 = “ block1_conv1 and block1_conv2”, b2-c1 and b2-c2 = “ block2_conv1 and block2_conv2”, b3-c1 and b3-c2 and b3-c3 and b3-c4 = ” block3_conv1 and block3_conv2 and block3_conv3 and block3_conv4”, b5-c1 and b5-c2 and b5-c3 and b5-c4 and b5-c5 = ” block5_conv1 and block5_conv2 and block5_conv3 and block5_conv4”. fla1 = ” flatten_1” t5 = 4608. dse3, dse4, dse5 = ” dense_3, dense_4, dense_5”, dt2 = ” droupout_2.