Open-vocabulary semantic segmentation on COCO-Stuff 164k
23.3mIoUSigLIPv2 + UniRefiner
Evaluation results
| Method | Links | |
|---|---|---|
| SigLIPv2 + UniRefinerViT=So/16, Background category=Without background category, Protocol=Zero-shot2026.05 | 23.3 | |
| EVA-CLIP + UniRefinerViT=8B/14, Background category=Without background category, Protocol=Zero-shot2026.05 | 20.5 | |
| CLIP + UniRefinerViT=G/14, Background category=Without background category, Protocol=Zero-shot2026.05 | 19.6 | |
| CLIPViT=G/14, Background category=Without background category, Protocol=Zero-shot2026.05 | 18.2 | |
| SigLIPv2ViT=So/16, Background category=Without background category, Protocol=Zero-shot2026.05 | 16.8 | |
| EVA-CLIPViT=8B/14, Background category=Without background category, Protocol=Zero-shot2026.05 | 10.8 |