Table 1 Comparison of computational complexity and segmentation performance between SS2D and KAN linear attention.
From: VMKLA-UNet: vision Mamba with KAN linear attention U-Net
Model | Parameters | FLOPs | Segmentation performance | ||
|---|---|---|---|---|---|
Dataset | mIoU (%) | DSC (%) | |||
SS2D | 0.06M | 161.219 M | ISIC17 | 80.23 | 89.03 |
ISIC18 | 81.35 | 89.71 | |||
KAN linear attention | 0.19 M | 3.539M | ISIC17 | 81.12 | 89.16 |
ISIC18 | 82.45 | 90.32 | |||
