Semantic Segmentation on Dataset 25% train data (test) benchmark leaderboard · SOTA2 Research

Semantic Segmentation on Dataset 25% train data (test)

70.2F1 Score (w/bg)FORTRESS

Evaluation Results

Method	Links
FORTRESS2026.01		70.2	66.9	56.8	52.3	76.9	67.6	55.4
SA-UNet2026.01		67	63.3	53.3	48.3	70.3	64.5	54.9
Att. U-Net2026.01		66.7	62.9	53.3	48.3	63.9	64.2	55.5
U-KAN2026.01		66.6	62.9	53.1	48.1	63.3	64.1	54
UNet3+2026.01		66	62.2	52.5	47.4	64.3	63.1	54.2
BiFPN2026.01		65.7	61.9	51.9	46.7	61.1	63.4	55.4
UNeXt2026.01		65.1	61.3	51.1	45.9	62.1	62.3	51.1
UNet++2026.01		63.2	59.1	49.7	44.3	56.8	61.1	51.9
U-Net2026.01		62.2	58	48.3	42.8	59.3	58.7	49.6
Swin-UNet2026.01		60.3	55.8	46.8	41.2	57.5	56.6	47.6
HierarchicalViT U-Net2026.01		56.7	51.9	43.5	37.5	60.8	53.2	41.3
FPN2026.01		55.4	50.4	41.9	35.8	60.2	53.2	42.1
EGE-UNet2026.01		45.8	39.6	34	27	40.9	42.4	36.6
MobileUNETR2026.01		13.3	3.2	11.4	1.9	13.1	5	5.8
FasterVit2026.01		12.7	2.6	11.1	1.5	12.5	4.5	4.2
Segformer2026.01		12.1	1.9	10.6	1	12.2	4.3	1.1
Rolling UNet-L2026.01		11.2	0.9	10.2	0.5	11.5	2.7	1.2