Unsupervised Semantic Segmentation With 10
Métriques
mIoU
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | mIoU |
|---|---|
| tagclip-a-local-to-global-framework-to | 35.3 |
| proxyclip-proxy-attention-improves-clip-for | 39.2 |
| tagalign-improving-vision-language-alignment | 33.3 |
| harnessing-vision-foundation-models-for-high | 42.2 |
| learning-to-generate-text-grounded-mask-for | 31.6 |
| cosmos-cross-modality-self-distillation-for | 31.3 |
| ttd-text-tag-self-distillation-enhancing | 37.4 |
| ttd-text-tag-self-distillation-enhancing | 26.5 |
| denseclip-extract-free-dense-labels-from-clip | 20.6 |
| groupvit-semantic-segmentation-emerges-from | 27.5 |
| reco-retrieve-and-co-segment-for-zero-shot-1 | 15.7 |