Retinexformer (https://arxiv.org/pdf/2303.06705.pdf) psnr result in table 1 is not consistent with Origin SID Paper's psnr result

[Li-wuhan]

In Retinexformer (https://arxiv.org/pdf/2303.06705.pdf), table 1 shows SID method only get PSNR 16.97 on SID Sony Dataset. But from 'Learn to see in the dark ' paper (https://arxiv.org/pdf/1805.01934.pdf)，SID method get PSNR 28.88 on SID Sony Dataset. Is the PSNR data (16.97) correct?

[caiyuanhao1998]

Hello, the subset is not exactly the same as the original SID paper. We copy this result from SNR-Net (https://openaccess.thecvf.com/content/CVPR2022/html/Xu_SNR-Aware_Low-Light_Image_Enhancement_CVPR_2022_paper.html)