.png){kind=icon}
.png){kind=icon}
Return
IGMamba: Inverse external with gated attention driven mamba for biomedical image segmentation
.png){kind=icon}
PRE
AI
C
Z
H
J
Z
S
李
李
肖
DOI：10.1016/j.eswa.2026.133837.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• IGMamba combines local feature extraction with long-range dependency modeling. • IEAD combines inverse external attention with multi-scale dilated convolutions. • Gated attention adaptively refines fused local-global features. • IGMamba achieves 92.35% DSC and 85.79% mIoU on the SCG dataset at over 30 FPS.
Keywords:
Mamba
Gated attention
Inverse external attention dilation
Biomedical image segmentation
Journal
IF:
7.5
Papers:
2.9W
Citations:
10.2W
Organization
Cited Papers