You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
By my understanding, most work train PSNR and MS-SSIM optimized models separately. Hence, I was wondering whether there is any reason for fine-tuning an MSE trained model with MS-SSIM rather than using MS-SSIM as an objective on a newly initialized model.
Did you find better results with the finetuning approach, or did you want to save on train time by re-using the PSNR optimized weights?
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
Hi,
By my understanding, most work train PSNR and MS-SSIM optimized models separately. Hence, I was wondering whether there is any reason for fine-tuning an MSE trained model with MS-SSIM rather than using MS-SSIM as an objective on a newly initialized model.
Did you find better results with the finetuning approach, or did you want to save on train time by re-using the PSNR optimized weights?
Beta Was this translation helpful? Give feedback.
All reactions