Zero Shot Semantic Segmentation On Pascal Voc
Métriques
Inductive Setting hIoU
Transductive Setting hIoU
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | Inductive Setting hIoU | Transductive Setting hIoU |
|---|---|---|
| exploring-regional-clues-in-clip-for-zero | 88.4 | 93.0 |
| denseclip-extract-free-dense-labels-from-clip | - | 87.4 |
| zegclip-towards-adapting-clip-for-zero-shot | 84.3 | 91.1 |
| zero-shot-semantic-segmentation-with | 80.8 | - |
| semantic-projection-network-for-zero-and-few | - | 38.8 |
| context-aware-feature-generation-for-zero | - | 43.7 |
| otseg-multi-prompt-sinkhorn-attention-for | 84.5 | 94.2 |
| freeseg-free-mask-from-interpretable | - | 86.9 |
| otseg-multi-prompt-sinkhorn-attention-for | 87.4 | 94.4 |
| 2112-14757 | 77.5 | 79.3 |
| decoupling-zero-shot-semantic-segmentation | 73.3 | - |
| 190600817 | - | 33.8 |
| a-closer-look-at-self-training-for-zero-label | - | 49.8 |