An Enhanced Technique of COVID-19 Detection and Classification Using Deep Convolutional Neural Network from Chest X-Ray and CT Images

Table 5

Performance comparisons between the IDConv-Net and state-of-the-art models on X-ray images.

Model	Dataset	Precision	Recall	-score	Training accuracy	Testing accuracy
AlexNet [57]	Lungs X-ray	69.25	90.43	81.83	69.84	67.76
nCOVnet [58]	Chest X-ray	82.00	97.62	89.13	97.00	88.10
Deep CNN [59]	covid-chestxray	99.17	71.76	83.27	72.78	71.90
InceptionResnetV2 [60]	covid-chestxray	92.11	92.38	92.07	93.83	92.18
MobileNetV2 [61]	covid-chestxray	20.00	100	33.33	62.12	60.00
ResNetV2 [61]	covid-chestxray	40.00	100	57.14	71.89	70.00
VGG-16 [62]	COVID-19 and pneumonia	86.17	86.23	86.38	87.36	86.39
AlexNet [40]	Covid_Data	84.62	94.57	89.32	92.93	82.62
DenseNet201 [61]	covid-chestxray	100	83.33	90.91	92.23	90.00
DenseNet121 [63]	Radiography Database	89.47	100	94.44	—	94.74
Proposed IDConv-Net model	Merged X-ray	97.14	91.87	94.43	97.49	96.99