From: Automated classification of polyps using deep learning architectures and few-shot learning
Model | Learning rate | Val-acc | Training speed | ||||
---|---|---|---|---|---|---|---|
 | 0.01 | 0.001 | 0.0001 | 0.00016 |  | Min/Epoch | #Epochs |
BiT-R152x4 | \(\checkmark\) | Â | Â | Â | 0.7890 | \(\approx\) 30 | 4 |
 |  | \(\checkmark\) |  |  | 0.8213 | \(\approx\) 30 | 8 |
 |  |  | \(\checkmark\) |  | 0.8140 | \(\approx\) 30 | 10 |
 |  |  |  | \(\checkmark\) | 0.8156 | \(\approx\) 30 | 10 |
EfficientNet-B7 | \(\checkmark\) | Â | Â | Â | 0.7903 | \(\approx\) 5.7 | 6 |
 |  | \(\checkmark\) |  |  | 0.8212 | \(\approx\) 5.7 | 10 |
 |  |  | \(\checkmark\) |  | 0.7924 | \(\approx\) 5.7 | 30 |
 |  |  |  | \(\checkmark\) | 0.7969 | \(\approx\) 5.7 | 28 |
Ours | \(\checkmark\) | Â | Â | Â | 0.4668 | \(\approx\) 3 | 19 |
 |  | \(\checkmark\) |  |  | 0.5938 | \(\approx\) 3 | 23 |
 |  |  | \(\checkmark\) |  | 0.8242 | \(\approx\) 3 | 10 |
 |  |  |  | \(\checkmark\) | 0.8950 | \(\approx\) 3 | 8 |