coincheung / mfm Goto Github PK

Hi, thanks for your great reproduction. Could you provide the current pretrained weights of your reproduction model?

Thanks

Hi,

I noticed that you removed the weight that balances the losses of different positions, which is a part of the focal frequency loss.

I wonder why that is. Did you observe a performance degradation?

I'm working on a similar project that involves spectrum reconstruction, and could use your advice. Intuitively, given the unbalanced distribution of the spectrum, the weight seems a good choice.

Thank you!

Hi, the weights of converting RGB image to gray image are 0.299, 0.587, and 0.114. But they are 0.229, 0.587, and 0.114 in your code. Perhaps that would be the potential reason. Thanks for your great work.