List your company
ResearchFormerly WizWand

Semantic Segmentation on COCO-Stuff without background class

48.6mIoUDouC
8.97619.26329.5539.837Dec 1, 2022Jun 26, 2023Jan 19, 2024Aug 13, 2024Mar 8, 2025Oct 1, 2025Apr 27, 2026
Updated 6mo ago

Evaluation results

MethodLinks
DouCBackbone=ViT-H, Training=Training-free, Instance Guidance=SAM22026.0448.6
DouCBackbone=ViT-L, Training=Training-free, Instance Guidance=SAM22026.0447.7
CorrCLIPBackbone=ViT-H, Training=Training-free, Instance Guidance=SAM22026.0446.9
CorrCLIPBackbone=ViT-B, Training=Training-free, Instance Guidance=SAM22026.0446.4
CorrCLIPBackbone=ViT-L, Training=Training-free, Instance Guidance=SAM22026.0446.1
DouCBackbone=ViT-B, Training=Training-free, Instance Guidance=SAM22026.0445.2
TridentBackbone=ViT-H, Training=Training-free2026.0444.3
CLIPerBackbone=ViT-L, Training=Training-free2026.0443.6
FreeDABackbone=ViT-L, Training=Training-free2026.0443.5
TridentBackbone=ViT-B, Training=Training-free2026.0442.2
SC-CLIPBackbone=ViT-L, Training=Training-free2026.0440.6
FSABackbone=ViT-H, Training=Training-free2026.0440.5
CASSBackbone=ViT-B, Training=Training-free2026.0440.2
SFPBackbone=ViT-B, Training=Training-free2026.0439.9
FSABackbone=ViT-B, Training=Training-free2026.0439.9
DIH-CLIPBackbone=ViT-B, Training=Training-free2026.0439.7
ProxyCLIPBackbone=ViT-H, Training=Training-free2026.0439.6
FSABackbone=ViT-L, Training=Training-free2026.0438.1
ProxyCLIPBackbone=ViT-L, Training=Training-free2026.0437.7
NACLIPBackbone=ViT-B, Training=Training-free2026.0435.2
ResCLIPBackbone=ViT-L, Training=Training-free2026.0434.5
SCLIPBackbone=ViT-B, Training=Training-free2026.0434.2
TagAlignPost-processing=false, Training Datasets=CC12M, Zero-shot evaluation=true2023.1225.8
TagAlignPost-processing=true, Training Datasets=CC12M, Zero-shot evaluation=true2023.1223.1
TCLEvaluation Mode=Zero-shot2022.1222.4
TCLPost-processing=false, Training Datasets=CC3&12M, Zero-shot evaluation=true2023.1222.4
TCLPost-processing=true, Training Datasets=CC3&12M, Zero-shot evaluation=true2023.1219.6
MaskCLIPRefinement=With heuristic refinement2022.1216.4
Dong et al.Post-processing=true, Training Datasets=LAION-20M, Zero-shot evaluation=true2023.1216.4
ReCoPost-processing=false, Training Datasets=ImageNet1K, Zero-shot evaluation=true2023.1216.3
SegCLIPPost-processing=true, Training Datasets=CC12M+COCO, Zero-shot evaluation=true2023.1216.1
GroupViTPost-processing=false, Training Datasets=CC12M+YFCC, Zero-shot evaluation=true2023.1215.4
GroupViT (RedCaps)Training Data=RedCaps + CC12M2022.1215.3
CoCuPost-processing=false, Training Datasets=CC3&12M+YFCC, Zero-shot evaluation=true2023.1215.2
Dong et al.Post-processing=false, Training Datasets=LAION-20M, Zero-shot evaluation=true2023.1215.1
CoCuPost-processing=true, Training Datasets=CC3&12M+YFCC, Zero-shot evaluation=true2023.1214.9
ReCoEvaluation Mode=Zero-shot2022.1214.8
ReCoPost-processing=true, Training Datasets=ImageNet1K, Zero-shot evaluation=true2023.1214.8
MaskCLIP (Baseline)Refinement=None2022.1214.7
GroupViT (YFCC)Training Data=YFCC + CC12M2022.1212.6
GroupViTPost-processing=true, Training Datasets=CC12M+YFCC, Zero-shot evaluation=true2023.1212.6
OVSegmentorPost-processing=true, Training Datasets=CC12M*, Zero-shot evaluation=true2023.1210.5