Thank you for your insightful work on AFUNet.
However, I have a question regarding the reported results. I tried to evaluate the testing outputs from the provided code using the HDR-VDP2 metric. I used the MATLAB version from this repository:
https://github.com/ZihengZZH/deghosting-hdr/tree/master/hdr-vdp-2
I set ppd = 50 and used the color encoding rgb-bt.709, but the resulting HDR-VDP2 score was higher than expected, around 67.6780.
May I ask whether there is an official or recommended script for computing the metrics consistently with those reported in the paper?
Thank you very much.
Thank you for your insightful work on AFUNet.
However, I have a question regarding the reported results. I tried to evaluate the testing outputs from the provided code using the HDR-VDP2 metric. I used the MATLAB version from this repository:
https://github.com/ZihengZZH/deghosting-hdr/tree/master/hdr-vdp-2
I set ppd = 50 and used the color encoding rgb-bt.709, but the resulting HDR-VDP2 score was higher than expected, around 67.6780.
May I ask whether there is an official or recommended script for computing the metrics consistently with those reported in the paper?
Thank you very much.