Table 1 the results of the quantitative comparison of the related methods.
From: Mural inpainting via two-stage generative adversarial network
Method | FID\(\downarrow\) | LIPS\(\downarrow\) | PSNR\(\uparrow\) | SSIM\(\uparrow\) | MSE\(\downarrow\) |
|---|---|---|---|---|---|
Edge Connect | 103.96 | 0.3813 | 17.21 | 0.5201 | 0.0188 |
CTSDG | 93.23 | 0.3763 | 18.22 | 0.5475 | 0.0168 |
AOT-GAN | 85.01 | 0.3237 | 18.29 | 0.5621 | 0.0167 |
Ref. 27 | 63.14 | 0.2439 | 20.38 | 0.6776 | 0.0110 |
MAT | 94.21 | 0.3998 | 17.16 | 0.4198 | 0.0290 |
STMA-GAN | 59.37 | 0.2136 | 21.31 | 0.7126 | 0.0085 |
