List your company
ResearchFormerly WizWand

Semantic Segmentation on Pascal VOC Priority protocol (1/16 labeled ratio)

87.3mIoUUniMatchV2-S + PrevMatch
63.1269.397575.67581.9525Oct 16, 2022Mar 30, 2023Sep 11, 2023Feb 24, 2024Aug 7, 2024Jan 19, 2025Jul 4, 2025
Updated 8mo ago

Evaluation results

MethodLinks
UniMatchV2-S + PrevMatchEncoder=DINOv2-S2024.0587.3
UniMatchV2-SEncoder=DINOv2-S2024.0586.6
SemiOVS (w/ PrevMatch)Encoder=R-1012025.0782
AllSparkEncoder=MiT-B52025.0781.6
AllSparkEncoder=MiT-B52024.0581.6
SemiOVS (w/ UniMatch)Encoder=R-1012025.0781.5
PrevMatchEncoder=R-1012025.0781.4
UniMatch + PrevMatchEncoder=RN-1012024.0581.4
CorrMatchEncoder=R-1012025.0781.3
CorrMatchEncoder=RN-1012024.0581.3
UniMatchEncoder=R-1012025.0780.9
UniMatchEncoder=RN-1012024.0580.9
Dual TeacherEncoder=R-1012025.0780.1
Dual TeacherEncoder=RN-1012024.0580.1
AugSegEncoder=R-1012025.0779.3
U2PLEncoder=R-1012025.0777.2
U2PLEncoder=RN-1012024.0577.2
U2PL+Encoder=RN-1012024.0577.2
FixMatch + PrevMatchEncoder=RN-502024.0576.9
FixMatchEncoder=RN-502024.0575.4
FPL+CPS w/ cutmixBackbone=ResNet-1012022.1074.98
FPL+AELBackbone=ResNet-1012022.1074.98
CPS w/ cutmixBackbone=ResNet-1012022.1074.48
AELBackbone=ResNet-1012022.1074.2
FPL+CPS w/o cutmixBackbone=ResNet-1012022.1073.18
CutMix-SegBackbone=ResNet-1012022.1072.56
FPL+CPS w/ cutmixBackbone=ResNet-502022.1072.52
CPS w/o cutmixBackbone=ResNet-1012022.1072.5
CACBackbone=ResNet-1012022.1072.4
CPS w/ cutmixBackbone=ResNet-502022.1071.78
FPL+AELBackbone=ResNet-502022.1071.01
Supervised BaselineEncoder=R-1012025.0770.6
MTBackbone=ResNet-1012022.1070.59
CACBackbone=ResNet-502022.1070.1
AELBackbone=ResNet-502022.1069.93
GCTBackbone=ResNet-1012022.1069.77
CutMix-SegBackbone=ResNet-502022.1068.9
FPL+CPS w/o cutmixBackbone=ResNet-502022.1068.67
CPS w/o cutmixBackbone=ResNet-502022.1068.13
CCTBackbone=ResNet-1012022.1067.94
MTBackbone=ResNet-502022.1066.77
CCTBackbone=ResNet-502022.1065.22
GCTBackbone=ResNet-502022.1064.05