DeepFillv2 (20-30% free form)

Reported on 4 benchmarks across 2 tasks · 1 paper · 2 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Medical2 results

Image GenerationonPlaces2 val
FID· 2018-06-10
13.5
best: 11.7 (Palatte (20-30% free form))
SOTA
Free-Form Image Inpainting with Gated Convolution arXiv:1806.03589
Image GenerationonPlaces2 val
PD· 2018-06-10
63
best: 115.4 (HiFill (128×128 center mask))
Free-Form Image Inpainting with Gated Convolution arXiv:1806.03589

Computer Vision2 results

Image InpaintingonPlaces2 val
FID· 2018-06-10
13.5
best: 11.7 (Palatte (20-30% free form))
SOTA
Free-Form Image Inpainting with Gated Convolution arXiv:1806.03589
Image InpaintingonPlaces2 val
PD· 2018-06-10
63
best: 115.4 (HiFill (128×128 center mask))
Free-Form Image Inpainting with Gated Convolution arXiv:1806.03589