List your company
ResearchFormerly WizWand

Semantic Segmentation on ADE-847

17.6mIoUCAT-Seg+DeCLIP
0.1284.6649.213.736Feb 23, 2023Aug 8, 2023Jan 22, 2024Jul 7, 2024Dec 21, 2024Jun 6, 2025Nov 20, 2025
Updated 7mo ago

Evaluation results

MethodLinks
CAT-Seg+DeCLIPBackbone=ViT-L/14, Training Set=COCO-Stuff2025.0517.6
CAT-SegBackbone=ViT-L/14, Training Set=COCO-Stuff2025.0516
CAT-Seg+DeCLIPBackbone=ViT-B/16, Training Set=COCO-Stuff2025.0515.3
FC-CLIPBackbone=ConvNeXt-L, Training Set=COCO-Panoptic2025.0514.8
FrozenSegBackbone=ConvNeXt-L, Training Set=COCO-Panoptic2025.0514.8
SAM 32025.1113.8
SAM 32025.1113.8
SANVL-Model=CLIP ViT-L/14, Training Dataset=COCO, ensemble=yes2023.0213.7
SANBackbone=ViT-L/14, Training Set=COCO-Stuff2025.0513.7
MAFTBackbone=ConvNeXt-L, Training Set=COCO-Stuff2025.0513.1
SANVL-Model=CLIP ViT-L/14, Training Dataset=COCO, ensemble=no2023.0212.4
SANSupervision Type=Fully-supervised, Annotation=true2024.0812.4
MAFTSupervision Type=Fully-supervised, Annotation=true2024.0812.1
CAT-SegBackbone=ViT-B/16, Training Set=COCO-Stuff2025.0512
ODISESupervision Type=Fully-supervised, Annotation=true2024.0811.1
ProxyCLIPSupervision Type=Training-free, Annotation=false2024.0811.1
ODISEBackbone=ViT-L/14, Training Set=COCO-Panoptic2025.0511.1
SANVL-Model=CLIP ViT-B/16, Training Dataset=COCO, ensemble=yes2023.0210.7
MasQCLIPSupervision Type=Fully-supervised, Annotation=true2024.0810.7
SANVL-Model=CLIP ViT-B/16, Training Dataset=COCO, ensemble=no2023.0210.1
HIPIESupervision Type=Fully-supervised, Annotation=true2024.089.7
X-DecoderSupervision Type=Fully-supervised, Annotation=true2024.089.2
APE-D*training=partially trained on LVIS2025.119.2
APE-D2025.119.2
OvSegVL-Model=CLIP ViT-L/14, Training Dataset=COCO, ensemble=yes2023.029
OVSegSupervision Type=Fully-supervised, Annotation=true2024.089
OVSegBackbone=ViT-L/14, Training Set=COCO-Stuff2025.059
OpenSegVL-Model=ALIGN EN-B7, Training Dataset=COCO+Loc. Narr., ensemble=no2023.028.8
MaskCLIPVL-Model=CLIP ViT-L/14, Training Dataset=COCO, ensemble=no2023.028.2
MaskCLIPSupervision Type=Fully-supervised, Annotation=true2024.088.2
OvSegVL-Model=CLIP ViT-B/16, Training Dataset=COCO, ensemble=yes2023.027.1
SimSeg (reproduced)VL-Model=CLIP ViT-L/14, Training Dataset=COCO, ensemble=yes2023.027.1
DeOPSupervision Type=Fully-supervised, Annotation=true2024.087.1
CLIP-DINOiserSupervision Type=Weakly-supervised, Annotation=false2024.087.1
SimSegVL-Model=CLIP ViT-B/16, Training Dataset=COCO, ensemble=yes2023.027
ZSsegBackbone=ViT-B/16, Training Set=COCO-Stuff2025.057
SimSeg (reproduced)VL-Model=CLIP ViT-B/16, Training Dataset=COCO, ensemble=yes2023.026.9
OpenSegVL-Model=ALIGN EN-B7, Training Dataset=COCO, ensemble=no2023.026.3
ZegFormer+Backbone=ViT-B/16, Training Set=COCO-Stuff2025.055.6
TCLSupervision Type=Weakly-supervised, Annotation=false2024.084.9
SCLIPSupervision Type=Training-free, Annotation=false2024.084.9
OpenSegVL-Model=ALIGN RN101, Training Dataset=COCO, ensemble=no2023.024
LSeg+VL-Model=ALIGN EN-B7, Training Dataset=COCO, ensemble=no2023.023.8
MaskCLIPSupervision Type=Training-free, Annotation=false2024.083.6
GKCSupervision Type=Fully-supervised, Annotation=true2024.083.5
LSeg+VL-Model=ALIGN RN101, Training Dataset=COCO, ensemble=no2023.022.5
CLIPSupervision Type=Training-free, Annotation=false2024.080.8