Table 5 Performance comparison of CancerDet-Net with baseline models.
From: Cross-platform multi-cancer histopathology classification using local-window vision transformers
Model | Test accuracy (%) | Recall (%) | F1 score (%) | Precision (%) |
|---|---|---|---|---|
InceptionV3 | 94.67 | 90.50 | 97.00 | 90.81 |
MobileNetV2 | 92.27 | 88.22 | 95.00 | 88.65 |
Xception | 95.55 | 95.55 | 95.55 | 95.55 |
VGG16 | 94.95 | 92.10 | 94.00 | 92.33 |
DenseNet121 | 92.90 | 91.20 | 92.30 | 91.10 |
ResNet50 | 92.50 | 90.80 | 92.00 | 90.75 |
EfficientNetB0 | 88.35 | 86.50 | 88.10 | 86.00 |
ShuffleNetV2 | 81.20 | 79.80 | 81.00 | 79.50 |
CancerDet-Net | 98.51 | 98.51 | 98.51 | 98.52 |
