Table 7 Results of comparison experiments about DCA module
From: DB-DCAFN: dual-branch deformable cross-attention fusion network for bacterial segmentation
Methods | Dice (mean ± SD, %) | |||
|---|---|---|---|---|
Aba | Kpn | Pae | All | |
DCA (apply offset points on local features) | 66.67 ± 12.79 | 65.43 ± 16.13 | 69.42 ± 14.66 | 67.23 ± 14.78 |
DCA (apply offset points on global features) | 70.57 ± 10.63 | 65.36 ± 20.62 | 70.49 ± 14.90 | 68.73 ± 16.26 |