Giters

yuanming-hu / exposure

Learning infinite-resolution image processing with GAN and RL from unpaired image datasets, using a differentiable photo editing model.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Some questions about mask

zhongtao93 opened this issue · comments

commented

Hi, thank you for you sharing code! I found that exposureNet can generate a mask when generate a filter parameters, however, only 5-6 parameters and get_mask() function can produce a HW1 mask. There is no hint in paper about how to train the mask and how to generate the mask, so can you help me to explain theory about the mask using in exposureNet.

Thank you very much!

commented

Hi,

Thanks for the question. Actually, the mask is an experimental feature and in the all results reported in the paper, masking is disabled.

Basically, you only need a few parameters for an alpha mask, whose alpha value is a linear combination of x, y, lumiance etc. Then we pass it to a sigmoid function to make its range within [0, 1].