List your company
ResearchFormerly WizWand

Open-Vocabulary Segmentation on Vaihingen (test)

63mIoUCONCEPTBANK
8.71222.80636.950.994Feb 6, 2026Feb 20, 2026Mar 7, 2026Mar 21, 2026Apr 5, 2026Apr 19, 2026May 4, 2026
Updated 6mo ago

Evaluation results

MethodLinks
CONCEPTBANKBackbone (Size)=SAM3 (PE-L+/14), Training Paradigm=Parameter-free, Evaluation Protocol=with background2026.0263
SegEarth-OV3Backbone (Size)=SAM3 (PE-L+/14), Training Paradigm=Parameter-free, Evaluation Protocol=with background, Pub. & Year=arXiv’252026.0260.8
SkySense-OBackbone (Size)=CLIP (ViT-L/14@512px), Training Paradigm=Training-based, Evaluation Protocol=with background, Pub. & Year=CVPR’252026.0251.6
SAM3Backbone (Size)=SAM3 (PE-L+/14), Training Paradigm=Parameter-free, Evaluation Protocol=with background, Pub. & Year=ICLR’262026.0249.1
ProxyCLIPBackbone (Size)=CLIP + DINOv2 (ViT-B/14), Training Paradigm=Parameter-free, Evaluation Protocol=with background, Pub. & Year=ECCV’242026.0247.5
CAFe-DINORS Training=No, Background Inclusion=Yes, Trained on target dataset=false2026.0547.1
CorrCLIPBackbone (Size)=CLIP + DINO (ViT-B/8) + SAM2 (Hiera-L), Training Paradigm=Parameter-free, Evaluation Protocol=with background, Pub. & Year=ICCV’252026.0247
RSKT-SegBackbone (Size)=CLIP (ViT-L/14@336px) + RemoteCLIP (ViT-B/16) + DINO (ViT-B/32), Training Paradigm=Training-based, Evaluation Protocol=with background, Pub. & Year=AAAI’262026.0242.7
Cat-SegBackbone (Size)=CLIP (ViT-L/14@336px), Training Paradigm=Training-based, Evaluation Protocol=with background, Pub. & Year=CVPR’242026.0242.3
SegEarth-OVBackbone (Size)=CLIP (ViT-B/16), Training Paradigm=Parameter-free, Evaluation Protocol=with background, Pub. & Year=CVPR’252026.0240
SANBackbone (Size)=CLIP (ViT-L/14@336px), Training Paradigm=Training-based, Evaluation Protocol=with background, Pub. & Year=CVPR’232026.0239.2
GEMBackbone (Size)=CLIP (ViT-B/16), Training Paradigm=Parameter-free, Evaluation Protocol=with background, Pub. & Year=CVPR’242026.0236.4
SCLIPBackbone (Size)=CLIP (ViT-B/16), Training Paradigm=Parameter-free, Evaluation Protocol=with background, Pub. & Year=ECCV’242026.0235.9
MaskCLIPBackbone (Size)=CLIP (ViT-B/16), Training Paradigm=Parameter-free, Evaluation Protocol=with background, Pub. & Year=ECCV’222026.0229.9
GSNetRS Training=Yes, Background Inclusion=Yes, Trained on target dataset=false2026.0528.2
SegEarth-OVRS Training=Self-Supervised, Background Inclusion=Yes, Trained on target dataset=false2026.0524.3
OVRSRS Training=Yes, Background Inclusion=Yes, Trained on target dataset=false2026.0519
DINOv3.txtRS Training=No, Background Inclusion=Yes, Trained on target dataset=false2026.0517.3
CLIPBackbone (Size)=CLIP (ViT-B/16), Training Paradigm=Parameter-free, Evaluation Protocol=with background, Pub. & Year=ICML’212026.0210.8