Unsupervised Semantic Segmentation With 3
評価指標
mIoU
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
比較表
| モデル名 | mIoU |
|---|---|
| learning-to-generate-text-grounded-mask-for | 24.0 |
| ttd-text-tag-self-distillation-enhancing | 32.0 |
| reco-retrieve-and-co-segment-for-zero-shot-1 | 24.2 |
| perceptual-grouping-in-vision-language-models | 18.1 |
| denseclip-extract-free-dense-labels-from-clip | 10.0 |
| harnessing-vision-foundation-models-for-high | 47.6 |
| reco-retrieve-and-co-segment-for-zero-shot-1 | 19.3 |
| ttd-text-tag-self-distillation-enhancing | 27.0 |
| cosmos-cross-modality-self-distillation-for | 34.7 |
| proxyclip-proxy-attention-improves-clip-for | 42.0 |
| tagalign-improving-vision-language-alignment | 27.5 |