Table 2. Evaluation results of the proposed model and [13], conducted on rendered calibration pattern images.
| Without illumination correction | With illumination correction |
| Ours | [13] | Ours | [13] |
| D-Gauss | PSNR ↑ | 35.296 ± 3.386 | 9.407 ± 2.585 | 36.634 ± 4.329 | 26.546 ± 4.500 |
| SSIM ↑ | 0.979 ± 0.021 | 0.361 ± 0.251 | 0.980 ± 0.021 | 0.829 ± 0.181 |
| LPIPS ↓ | 0.146 ± 0.018 | 0.496 ± 0.301 | 0.143 ± 0.017 | 0.095 ± 0.070 |
| Cooke | PSNR ↑ | 34.754 ± 3.571 | 8.978 ± 2.392 | 35.533 ± 4.262 | 28.101 ± 4.862 |
| SSIM ↑ | 0.973 ± 0.035 | 0.361 ± 0.249 | 0.973 ± 0.035 | 0.873 ± 0.174 |
| LPIPS ↓ | 0.020 ± 0.028 | 0.471 ± 0.304 | 0.020 ± 0.028 | 0.064 ± 0.064 |
| Tessar | PSNR ↑ | 26.034 ± 5.018 | 8.224 ± 2.602 | 26.281 ± 5.218 | 27.185 ± 4.661 |
| SSIM ↑ | 0.805 ± 0.115 | 0.365 ± 0.253 | 0.805 ± 0.115 | 0.872 ± 0.166 |
| LPIPS ↓ | 0.142 ± 0.106 | 0.455 ± 0.289 | 0.141 ± 0.106 | 0.065 ± 0.073 |