Zero Shot Semantic Segmentation On Coco Stuff
Metrics
Inductive Setting hIoU
Transductive Setting hIoU
Results
Performance results of various models on this benchmark
Comparison Table
| Model Name | Inductive Setting hIoU | Transductive Setting hIoU |
|---|---|---|
| zegclip-towards-adapting-clip-for-zero-shot | 40.8 | 48.5 |
| zero-shot-semantic-segmentation-with | 38.2 | - |
| context-aware-feature-generation-for-zero | 18.2 | 19.5 |
| a-closer-look-at-self-training-for-zero-label | - | 34.8 |
| otseg-multi-prompt-sinkhorn-attention-for | 41.4 | 49.5 |
| semantic-projection-network-for-zero-and-few | 14.0 | 30.3 |
| denseclip-extract-free-dense-labels-from-clip | - | 45.0 |
| exploring-regional-clues-in-clip-for-zero | 41.2 | 49.7 |
| mvp-seg-multi-view-prompt-learning-for-open | - | 45.5 |
| 2112-14757 | 36.3 | 41.5 |
| decoupling-zero-shot-semantic-segmentation | 33.2 | - |
| otseg-multi-prompt-sinkhorn-attention-for | 41.5 | 49.8 |
| freeseg-free-mask-from-interpretable | - | 45.3 |
| sign-spatial-information-incorporated | 20.9 | - |
| 190600817 | 15.0 | 16.2 |