List your company
ResearchFormerly WizWand

Semantic Segmentation on MultiSenGe 0.3 size (train)

44.75mIoULandSegmenter
28.390832.637936.88541.1321Nov 11, 2025
Updated 6mo ago

Evaluation results

MethodLinks
LandSegmenterBackbone=SAM2+DOFA, Evaluation Protocol=Fine-tuned2025.1144.75
SAM2+HR+DOFA*Backbone=SAM2+DOFA, Evaluation Protocol=Fine-tuned2025.1142.12
SAM2+HR (3B)Backbone=SAM2, Parameter Count=3B, Evaluation Protocol=Fine-tuned2025.1140.58
DOFABackbone=ViT-Large, Segmentation Framework=UperNet, Evaluation Protocol=Fine-tuned2025.1140.1
scaleMAE (3B)Backbone=ViT-Large, Segmentation Framework=UperNet, Parameter Count=3B, Evaluation Protocol=Fine-tuned2025.1139.82
satMAE (3B)Backbone=ViT-Large, Segmentation Framework=UperNet, Parameter Count=3B, Evaluation Protocol=Fine-tuned2025.1137.76
CromSS (9B)Backbone=ResNet50, Segmentation Framework=DeepLabv3+, Parameter Count=9B, Evaluation Protocol=Fine-tuned2025.1136.37
satMAE++ (10B)Backbone=ViT-Large, Segmentation Framework=UperNet, Parameter Count=10B, Evaluation Protocol=Fine-tuned2025.1136.17
satMAE (10B)Backbone=ViT-Large, Segmentation Framework=UperNet, Parameter Count=10B, Evaluation Protocol=Fine-tuned2025.1136.15
satMAE++ (3B)Backbone=ViT-Large, Segmentation Framework=UperNet, Parameter Count=3B, Evaluation Protocol=Fine-tuned2025.1134.17
SAM2 (3B)Backbone=SAM2, Parameter Count=3B, Evaluation Protocol=Fine-tuned2025.1129.02