Image Inpainting On Places2 Val
Metrics
FID
PD
Results
Performance results of various models on this benchmark
Model Name | FID | PD | Paper Title | Repository |
|---|---|---|---|---|
| TFill (20-30% free form) | 15.2 | 87.2 | Bridging Global Context Interactions for High-Fidelity Image Completion | |
| HiFill (128×128 center mask) | 16.9 | 115.4 | Contextual Residual Aggregation for Ultra High-Resolution Image Inpainting | |
| DeepFillv2 (128×128 center mask) | 15.3 | 96.3 | Free-Form Image Inpainting with Gated Convolution | |
| DeepFillv2 (20-30% free form) | 13.5 | 63.0 | Free-Form Image Inpainting with Gated Convolution | |
| Palatte (20-30% free form) | 11.7 | 35.0 | Palette: Image-to-Image Diffusion Models | |
| HiFill (20-30% free form) | 15.7 | 92.8 | Contextual Residual Aggregation for Ultra High-Resolution Image Inpainting | |
| Palette (128×128 center mask) | 11.9 | 57.3 | Palette: Image-to-Image Diffusion Models |
0 of 7 row(s) selected.