Lightweight U-Net face super-resolution from 16x16 to 128x128 uses YOLO-World landmark heatmaps converted to spatial loss weights, improving metrics on CelebA without adversarial training or extra networks.
The unreasonable effectiveness of deep features as a perceptual metric
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
You Only Landmark Once: Lightweight U-Net Face Super Resolution with YOLO-World Landmark Heatmaps
Lightweight U-Net face super-resolution from 16x16 to 128x128 uses YOLO-World landmark heatmaps converted to spatial loss weights, improving metrics on CelebA without adversarial training or extra networks.