Skip to main content
. 2021 Sep 3;71:103126. doi: 10.1016/j.bspc.2021.103126

Table 2.

Architecture of the proposed model.

Layer(type) Output shape # of parameter
VGG(model) 512 20,025,923
GoogLeNet(model) 2048 21,802,784
DenseNet(model) 1920 18,321,984
NASNet(model) 1056 4,269,716
concatenate_6 (Concatenate) 5536 0
dropout(Dropout) 5536 0
dense(Dense) 256 1,417,216
dense_1(Dense) 3 768

Total Parameters 65,838,400
Parameters trained 1,417,985