Skip to main content
. 2024 Feb 27;10:e1862. doi: 10.7717/peerj-cs.1862

Table 4. Performance metrics (%) of pre-trained deep networks as a feature extractor.

Models Accuracy Sensitivity Specificity Precision F1-score
DenseNet201 9,707 9,570 9,841 9,625 9,597
DarkNet53 9,432 9,409 9,694 9,315 9,360
Xception 9,585 9,603 9,771 9,539 9,571