Skip to main content
. 2021 Sep 28;7:e715. doi: 10.7717/peerj-cs.715

Table 3. Results of the DUF classification task trained with 1,000 images.

Model Accuracy (%) Specificity (%) Recall (%) Precision (%) F1-Score (%)
Proposed model 87.6 86.2 88.8 87.3 88.5
Xception 87.4 86.3 88.4 87.3 87.9
ResNet50 88.6 86.4 90.9 87.4 89.0