Skip to main content
. 2021 Sep 27;18(19):10147. doi: 10.3390/ijerph181910147

Table 3.

Overall performance comparison of 3-class and binary classifier InceptionV3, ResNet50, and VGG19 models.

Model Classes Mean Accuracy Precision Recall Mean AUC
Inception V3 COVID-19, Normal 1 1 1 1
COVID-19, Normal, Pneumonia 97.10% 0.9713 0.9710 0.9784
ResNet50 COVID-19, Normal 1 1 1 1
COVID-19, Normal, Pneumonia 98.55% 0.9855 0.9855 0.9890
VGG19 COVID-19, Normal 1 1 1 1
COVID-19, Normal, Pneumonia 98.55% 0.9855 0.9855 0.9893