List your company
ResearchFormerly WizWand

Semantic Segmentation on Pascal Context 59 (val)

59.4mIoUMAFT+
20.9230.9140.950.89Dec 4, 2023Feb 3, 2024Apr 4, 2024Jun 4, 2024Aug 4, 2024Oct 4, 2024Dec 5, 2024
Updated 8mo ago

Evaluation results

MethodLinks
MAFT+VLM Backbone=ConvNeXt-L, Ensemble Operation=false2024.0859.4
SCANVL-Model=CLIP VIT-L/14, Training Dataset=COCO2023.1259.3
MAFTVLM Backbone=ViT-L, Ensemble Operation=false2024.0859
SCANVL-Model=CLIP VIT-B/16, Training Dataset=COCO2023.1258.4
FC-CLIPVLM Backbone=ConvNeXt-L, Ensemble Operation=true2024.0858.4
SANVL-Model=CLIP VIT-L/14, Training Dataset=COCO2023.1257.7
SANVLM Backbone=ViT-L, Ensemble Operation=false2024.0857.7
MAFTVLM Backbone=ConvNeXt-L, Ensemble Operation=false2024.0857.5
ODISEVL-Model=CLIP VIT-L/14, Training Dataset=COCO2023.1257.3
ODISEVLM Backbone=ViT-L, Ensemble Operation=false2024.0857.3
UniLSeg-100Additional LMM=false2023.1256.7
OVSegVL-Model=CLIP VIT-L/14, Training Dataset=COCO2023.1255.7
OVSegVLM Backbone=ViT-L, Ensemble Operation=false2024.0855.7
UniLSeg-20Additional LMM=false2023.1254.3
SANVL-Model=CLIP VIT-B/16, Training Dataset=COCO2023.1253.8
MAFTVL-Model=CLIP ViT-B/16, Training Dataset=COCO2023.1253.5
SAM w/ MAFTBackbone=SAM-H, Setting=Open-vocabulary2024.1253.5
OVSegVL-Model=CLIP VIT-B/16, Training Dataset=COCO2023.1253.3
OVSegAdditional LMM=true2023.1253.3
SimSeg†VL-Model=CLIP VIT-L/14, Training Dataset=COCO2023.1252.2
SimSeg†VL-Model=CLIP VIT-B/16, Training Dataset=COCO2023.1251.9
SAM w/ Mask-AdapterBackbone=SAM-H, Setting=Open-vocabulary2024.1249.5
OpenSegVL-Model=ALIGN EN-B7, Training Dataset=COCO+Loc. Narr.2023.1248.2
OpenSegVLM Backbone=ALIGN, Ensemble Operation=false2024.0848.2
SimSegVL-Model=CLIP VIT-B/16, Training Dataset=COCO2023.1247.7
SimSegAdditional LMM=true2023.1247.7
LSeg+VL-Model=ALIGN EN-B7, Training Dataset=COCO2023.1246.5
MaskCLIPVL-Model=CLIP VIT-L/14, Training Dataset=COCO2023.1245.9
MaskCLIPAdditional LMM=true2023.1245.9
GKCAdditional LMM=false2023.1245.2
FC-CLIPVLM Backbone=ConvNeXt-L, Ensemble Operation=false2024.0842.4
OpenSegVL-Model=ALIGN EN-B7, Training Dataset=COCO2023.1242.1
OpenSegAdditional LMM=true2023.1242.1
OpenSegVL-Model=ALIGN RN101, Training Dataset=COCO2023.1236.9
LSeg+VL-Model=ALIGN RN101, Training Dataset=COCO2023.1236
LSeg+Additional LMM=false2023.1236
SAM w/ CLIPBackbone=SAM-H, Setting=Open-vocabulary2024.1234.4
Group-VITVL-Model=rand. init., Training Dataset=CC12M+YFCC2023.1222.4