Update README.md

README.md

@@ -68,6 +68,4 @@ And lastly, I also ran multiple non-reference quality metrics on my set, and as
 <figure>
   <img src="https://cdn-uploads.huggingface.co/production/uploads/634e9aa407e669188d3912f9/IhobStl9Aj6fT0qU4728W.png" alt="Multiple non-reference set metrics">
   <figcaption>Multiple non-reference set metrics</figcaption>
-</figure> 
-
-(PS something I had also noticed in general, is that in papers mostly only psnr and ssim is used for metrics. But, a problem is, that psnr can also be run with y-channel true, which would give slightly higher metric numbers. And while I hope that this would be indicated in papers with "psnry" instead of "psnr" for example, I am not sure this is always the case. Also we have way more FR metrics available than those two)
+</figure>