Table 4.

Performance evaluation of different deep learning techniques on the multi-class dataset

State-of-the-art	Accuracy (%)
Dark COVID-Net (Ozturk et al. 2020)	87.02
MobileNetV2 (Apostolopoulos and Mpesiana 2020)	94.72
CORO-Net (Khan et al. 2020)	95
CNN + LSTM (Islam et al. 2020)	97
Covid-Net (Wang and Wong 2020)	93.3
Proposed ensemble model	99.21

Accuracy of the proposed ensemble model is highlighted to emphasize on the highest accuracy value with respect to the other techniques that have been utilized for three-class classification