Table 3A.
The comparison of the DRIVE data set's segmentation results using various segmentation techniques.
| Method | Year | Se | Sp | Acc |
|---|---|---|---|---|
| Cross modality learning (17) | 2015 | 0.7569 | 0.9816 | 0.9527 |
| GMM classifier (18) | 2015 | 0.7249 | 0.9830 | 0.9620 |
| SP model (19) | 2016 | 0.7811 | 0.9807 | 0.9535 |
| CRF model (20) | 2016 | 0.7897 | 0.9684 | – |
| VS method (21) | 2017 | 0.7779 | 0.9780 | 0.9521 |
| RU-Net and R2U-Net (9) | 2018 | 0.7792 | 0.9813 | 0.9556 |
| LadderNet (22) | 2018 | 0.7856 | 0.9810 | 0.9561 |
| U-Net+joint losses (23) | 2018 | 0.7653 | 0.9818 | 0.9542 |
| CTF-Net (24) | 2018 | 0.7979 | 0.9857 | 0.9685 |
| Three-stage DL Model (25) | 2019 | 0.7631 | 0.9820 | 0.9538 |
| SD-Unet (32) | 2019 | 0.7891 | 0.9848 | 0.9674 |
| Dilated Conv. (33) | 2019 | 0.7903 | 0.9813 | 0.9567 |
| GFM (15) | 2020 | 0.7614 | 0.9837 | 0.9604 |
| DL methods (10) | 2020 | 0.7979 | 0.9794 | 0.9563 |
| AA-UNet (34) | 2020 | 0.7941 | 0.9798 | 0.9558 |
| EDC-Net (35) | 2020 | 0.7092 | 0.9820 | 0.9447 |
| Iternet (36) | 2020 | 0.7735 | 0.9838 | 0.9673 |
| MLC scheme (37) | 2021 | 0.7761 | 0.9792 | 0.9519 |
| LAC network (38) | 2021 | 0.7921 | 0.9810 | 0.9568 |
| ResDo-UNet (39) | 2021 | 0.7985 | 0.9791 | 0.9561 |
| FPM-Net (proposed) | 2022 | 0.8285 | 0.98270 | 0.96920 |
“–” means the value is not available in the relevant research study.