Table 1.

Classification performance (%) of the examined deep neural network architectures following the baseline approach.

Base Model	Accuracy	F1-Score	Precision	Recall	Jaccard	Dice
EfficientNetB4	87.50	86.32	93.57	81.48	75.93	86.32
EfficientNetB7	83.75	76.39	91.45	70.86	69.40	81.94
VGG16	92.22	91.94	93.97	90.13	83.21	90.84
Xception	86.02	85.26	92.63	80.18	72.59	84.12
InceptionResNetV2	86.51	86.53	91.34	82.79	73.06	84.43
InceptionV3	86.82	84.79	90.73	80.58	72.60	84.12
MobileNetV2	86.68	85.76	91.87	81.33	72.76	84.24
ResNet50V2	89.02	88.90	93.56	85.14	77.11	87.08
DenseNet121	84.27	80.20	93.77	73.65	70.97	83.02
DenseNet169	87.23	86.52	92.84	82.03	74.75	85.55
DenseNet201	86.37	87.13	93.04	82.83	72.75	84.23

Note: Results in bold denote the best performance for each metric.