Open sp-1234 opened 6 years ago
I've done that with an earlier version of the library #15
For butteraugli this benchmark is problematic, because it also includes heavily distorted images, and butteraugli claims to only support comparison of very similar images.
oops, I somehow overlooked that…
maybe it could make sense to exclude images which are too dissimilar, and look at the correlation for "good" images.
http://www.ponomarenko.info/tid2013.htm Metrics are ranked by their Spearman correlation with Mean Opinion Score.
Someone who has hardware powerful enough to make it practical could do this. It would be a real proof (or maybe disproof) that this metric is superior to others.