Open-Vocabulary Segmentation on Pascal VOC 21 2012 (val)
87.1mIoUCONCEPTBANK
Evaluation Results
| Method | Links | |
|---|---|---|
| CONCEPTBANKBackbone (Size)=SAM3 (PE-L+/14), Optimization protocol=Parameter-free2026.02 | 87.1 | |
| ReMEBackbone (Size)=CLIP + DINOv2 (ViT-L/14) + SAM (ViT-L/16), Optimization protocol=Parameter-free2026.02 | 82.2 | |
| SAM3Backbone (Size)=SAM3 (PE-L+/14), Optimization protocol=Parameter-free2026.02 | 81.9 | |
| CorrCLIPBackbone (Size)=CLIP (ViT-L/14) + DINO (ViT-B/8) + SAM2 (Hiera-L), Optimization protocol=Parameter-free2026.02 | 76.7 | |
| RF-CLIPBackbone (Size)=CLIP (ViT-B/16), Optimization protocol=Parameter-free2026.02 | 67.2 | |
| TridentBackbone (Size)=CLIP + DINO (ViT-B/16) + SAM (ViT-B/16), Optimization protocol=Parameter-free2026.02 | 67.1 | |
| Talk2DINOBackbone (Size)=CLIP + DINOv2 with registers (ViT-B/14), Optimization protocol=Training-based2026.02 | 65.8 | |
| CASSBackbone (Size)=CLIP + DINO (ViT-B/8), Optimization protocol=Parameter-free2026.02 | 65.8 | |
| DIH-CLIPBackbone (Size)=CLIP (ViT-B/16), Optimization protocol=Parameter-free2026.02 | 64.2 | |
| SFPBackbone (Size)=CLIP (ViT-B/16), Optimization protocol=Parameter-free2026.02 | 63.9 | |
| CLIP-DINOiserBackbone (Size)=CLIP + DINO (ViT-B/16), Optimization protocol=Training-based2026.02 | 62.1 | |
| LaVGBackbone (Size)=CLIP + DINO (ViT-B/8), Optimization protocol=Parameter-free2026.02 | 62.1 | |
| LPOSSBackbone (Size)=CLIP + DINO (ViT-B/16), Optimization protocol=Parameter-free2026.02 | 61.1 | |
| SAM-CLIPBackbone (Size)=CLIP + SAM (ViT-B/16), Optimization protocol=Training-based2026.02 | 60.6 | |
| SCLIPBackbone (Size)=CLIP (ViT-B/16), Optimization protocol=Parameter-free2026.02 | 59.1 | |
| NACLIPBackbone (Size)=CLIP (ViT-B/16), Optimization protocol=Parameter-free2026.02 | 58.9 | |
| ProxyCLIPBackbone (Size)=CLIP + DINOv2 with registers (ViT-B/14), Optimization protocol=Parameter-free2026.02 | 58.6 | |
| CoDeBackbone (Size)=CLIP (ViT-B/16), Optimization protocol=Training-based2026.02 | 57.7 | |
| SegCLIPBackbone (Size)=CLIP (ViT-B/16), Optimization protocol=Training-based2026.02 | 52.6 | |
| FreeDABackbone (Size)=CLIP + DINOv2 (ViT-B/14), Optimization protocol=Parameter-free2026.02 | 51.8 | |
| ClearCLIPBackbone (Size)=CLIP (ViT-B/16), Optimization protocol=Parameter-free2026.02 | 51.8 | |
| TCLBackbone (Size)=CLIP (ViT-B/16), Optimization protocol=Training-based2026.02 | 51.2 | |
| CLIPtraseBackbone (Size)=CLIP (ViT-B/16), Optimization protocol=Parameter-free2026.02 | 50.9 | |
| GroupViTBackbone (Size)=GroupViT (ViT-S/16), Optimization protocol=Training-based2026.02 | 50.4 | |
| CaRBackbone (Size)=CLIP (ViT-B/16), Optimization protocol=Parameter-free2026.02 | 48.6 | |
| GEMBackbone (Size)=CLIP (ViT-B/16), Optimization protocol=Parameter-free2026.02 | 46.2 | |
| CLIPBackbone (Size)=CLIP (ViT-B/16), Optimization protocol=Parameter-free2026.02 | 18.6 |