Image To Image Translation On Ade20K Labels
評価指標
FID
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 | FID | mIoU |
|---|---|---|
| wavelet-based-unsupervised-label-to-image-1 | 34.5 | 16.95 |
| photographic-image-synthesis-with-cascaded | 73.3 | 22.4 |
| cross-domain-correspondence-learning-for | 26.4 | - |
| diverse-semantic-image-synthesis-via | - | - |
| you-only-need-adversarial-supervision-for-1 | 28.3 | 48.8 |
| dual-pyramid-generative-adversarial-networks | 26.1 | 52.7 |
| high-resolution-image-synthesis-and-semantic | 81.8 | 20.3 |
| unlocking-pre-trained-image-backbones-for | 22.7 | 54.3 |
| semantic-image-synthesis-with-spatially | 33.9 | 38.5 |
| taming-transformers-for-high-resolution-image | 35.5 | - |
| sesame-semantic-editing-of-scenes-by-adding | 31.9 | 49 |
| improving-augmentation-and-evaluation-schemes | 41.5 | - |
| improving-augmentation-and-evaluation-schemes | 32.6 | 44 |
| learning-to-predict-layout-to-image | 31.7 | 43.7 |
| stochastic-conditional-diffusion-models-for | 26.9 | 49.4 |
| usis-unsupervised-semantic-image-synthesis | 33.2 | 17.38 |