List your company
ResearchFormerly WizWand

Semantic Segmentation on PASCAL-Context (mIoU)

84.58mIoUTIGER-L
7.432827.461447.4967.5186Dec 2, 2021Sep 3, 2022Jun 6, 2023Mar 8, 2024Dec 9, 2024Sep 11, 2025Jun 14, 2026
Updated 4mo ago

Evaluation results

MethodLinks
TIGER-LBackbone=ViT-L2026.0684.58
TIGER-BBackbone=ViT-B2026.0684.08
SAK2026.0684.01
MTMamba++Decoder Type=Mamba-based decoder2026.0581.94
SEM2026.0681.66
TSP-TransformerDecoder Type=Transformer-based decoder2026.0581.48
MLoREDecoder Type=Transformer-based decoder2026.0581.41
MLoRE2026.0681.41
TaskDiffusionDecoder Type=Diffusion-based decoder2026.0581.21
MTMambaDecoder Type=Mamba-based decoder2026.0581.11
MTMamba2026.0681.11
RADIO2026.0681.11
DINOv2-ViT-LBackbone=ViT-L2026.0681.04
TaskPrompterDecoder Type=Transformer-based decoder2026.0580.89
TaskPrompter2026.0680.89
MTLSI-NetParams (M)=234, FLOPs (G)=7722026.0480.86
B^3-NetDecoder Type=Bridge-feature-based decoder2026.0580.81
TaskExpertParams (M)=420, FLOPs (G)=6222026.0480.64
TaskExpert2026.0680.64
InvPT++Params (M)=421, FLOPs (G)=6672026.0480.22
InvPT++Decoder Type=Transformer-based decoder2026.0580.22
BridgeNetDecoder Type=Bridge-feature-based decoder2026.0579.89
InvPTDecoder Type=Transformer-based decoder2026.0579.03
InvPT2026.0679.03
MQTransformer2026.0478.93
MQTransformerDecoder Type=Transformer-based decoder2026.0578.93
MQTransformer2026.0678.93
MTI-NetParams (M)=851, FLOPs (G)=7742026.0478.31
MTI-Net2026.0678.31
PAD-NetParams (M)=330, FLOPs (G)=7732026.0478.01
PAD-Net2026.0678.01
CLIP-ViT-LBackbone=ViT-L2026.0677.75
ATRCParams (M)=340, FLOPs (G)=8712026.0477.11
ATRC2026.0677.11
OWLv2-ViT-LBackbone=ViT-L2026.0673.22
ASTMTDecoder Type=CNN-based decoder2026.0568
ATRC-BMTASDecoder Type=CNN-based decoder2026.0567.67
ATRC-ASPPDecoder Type=CNN-based decoder2026.0563.6
ATRCDecoder Type=CNN-based decoder2026.0562.69
DiffusionMTL2026.0662.55
MTI-NetDecoder Type=CNN-based decoder2026.0561.7
SAM-ViT-BBackbone=ViT-B2026.0654.88
PAD-NetDecoder Type=CNN-based decoder2026.0553.6
DPFBackbone=ViT2023.0345.31
MaskCLIP+paradigm=transductive2021.1244.4
Fully Sup.2021.1244.4
Zhao et al.rGMM=true2023.0336.07
DeiTArch.=ViT, Init.=false, Training Data=ImageNet, Sup.=Class, Zero-Shot=false2022.1135.9
Zhao et al.rGMM=false2023.0333.54
Qian et al.Online Extension=true2023.0330
Qian et al.Online Extension=false2023.0329.7
ZS3Netparadigm=transductive2021.1227
CaGNetparadigm=inductive2021.1224.8
SegCLIPArch.=ViT, Init.=true, Training Data=CC+COCO, Sup.=Text, Zero-Shot=true2022.1124.7
MoCoArch.=ViT, Init.=false, Training Data=CC12M+YFCC, Sup.=Self, Zero-Shot=false2022.1123
DINOArch.=ViT, Init.=false, Training Data=CC12M+YFCC, Sup.=Self, Zero-Shot=false2022.1122.8
GroupViTArch.=ViT, Init.=false, Training Data=CC12M+YFCC, Sup.=Text, Zero-Shot=true2022.1122.4
ZS3Netparadigm=inductive2021.1220.8
SegCLIPArch.=ViT, Init.=false, Training Data=CC+COCO, Sup.=Text, Zero-Shot=true2022.1119.1
GroupViT1-sArch.=ViT, Init.=false, Training Data=CC+COCO, Sup.=Text, Zero-Shot=true, Stage=one-stage2022.1114.8
GroupViT2-sArch.=ViT, Init.=false, Training Data=CC+COCO, Sup.=Text, Zero-Shot=true, Stage=two-stage2022.1110.4