Table 3 Comparison of mean test scores by the best ResNet50 backbone LSTM model against a 3DResNet model without temporal analysis (both pretrained and trained from scratch).
From: Automatic classification of normal and abnormal cell division using deep learning
Architecture | Mean accuracy | Mean precision | Mean recall | Mean F1-score |
|---|---|---|---|---|
Best ResNet50 LSTM | 0.94 | 0.97 | 0.9 | 0.93 |
3D ResNet50 (from scratch) | 0.479 | 0.489 | 0.957 | 0.647 |
3D ResNet50 (pretrained) | 0.468 | 0.483 | 0.936 | 0.638 |
