Table 3.

Comparing of the proposed model's results with the state-of-the-art models’ results.

Authors	Approaches	Results
Wang et al. [2]	COVID-Net	ACC = 83.5% for 4-classes, ACC = 92.4% for 3-classes.
Hemdan et al. [3]	COVIDX-Net	F1-scores = 0.91 and 0.89 of 2-classed (i.e., COVID-19 and normal cases)
Sethy and Behera[4]	ResNet50+SVM	ACC = 95.33%
Ozturk et al. [5]	DarkNet	ACC = 87.02% of multi-class ACC = 98.08% of 2-classes.
Khan et al. [6]	]oroNet	ACC = 89.6% and 95% for 4 classes.
The proposed model	CycleGAN-Inception	ACC = 94.2% and AUC = 92.2%