Table 2 Comparisons with the state-of-the-art baselines on the CVC-ClinicDB dataset
Method | Year | mDice | mIoU | Recall | Precision |
|---|---|---|---|---|---|
FCN [45] | 2017 | 0.7732 | 0.8999 | ||
CNN [46] | 2018 | 0.87 | – | – | – |
SegNet [47] | 2018 | – | – | 0.8824 | – |
U-Net [5] | 2015 | 0.8781 | 0.7881 | 0.7865 | 0.9329 |
ResUNet++ [43] | 2019 | 0.9199 | 0.8892 | 0.9391 | 0.8445 |
DoubleU-Net [25] | 2020 | 0.9239 | 0.8611 | 0.8457 | 0.9592 |
TransUNet [12] | 2021 | 0.9350 | 0.8870 | – | – |
DS-TransUNet-B [14] | 2021 | 0.9350 | 0.8845 | 0.9464 | 0.9306 |
DS-TransUNet-L [14] | 2021 | 0.9422 | 0.8939 | 0.9500 | 0.9369 |
MSRF-Net [41] | 2021 | 0.9420 | 0.9043 | 0.9567 | 0.9427 |
EG-TransUNet | – | 0.9523 | 0.9130 | 0.9540 | 0.9536 |