Image Harmonization On Iharmony4
Métriques
MSE
PSNR
fMSE
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | MSE | PSNR | fMSE |
|---|---|---|---|
| region-aware-adaptive-instance-normalization | 40.29 | 36.12 | 469.60 |
| deep-image-harmonization-via-domain | 52.33 | 34.76 | 549.96 |
| intrinsic-image-harmonization | 38.71 | 35.90 | 400.29 |
| dccf-deep-comprehensible-color-filter | 24.65 | 37.87 | - |
| pct-net-full-resolution-image-harmonization | 18.80 | 39.28 | 238.27 |
| dense-pixel-to-pixel-harmonization-via | 24.62 | 38.07 | 296.31 |
| deep-image-harmonization-in-dual-color-spaces | 18.47 | 39.17 | 212.53 |
| phnet-patch-based-normalization-for-portrait | 16.05 | 38.54 | 178.11 |
| hierarchical-dynamic-image-harmonization | 16.55 | 40.46 | 179.49 |
| bargainnet-background-guided-domain | 37.82 | 35.88 | 405.23 |
| learning-global-aware-kernel-for-image | 19.90 | 39.53 | 220.44 |
| image-harmonization-with-transformer | 30.30 | 37.55 | 320.78 |
| harmonizer-learning-to-perform-white-box | 24.26 | 37.84 | 339.23 |
| spatial-separated-curve-rendering-network-for | 35.58 | 37.18 | 274.99 |
| hierarchical-dynamic-image-harmonization | 24.99 | 38.63 | 260.65 |
| high-resolution-image-harmonization-via | 23.75 | 38.23 | - |