Learning-to-See-in-the-Dark apply whitebalance before feed into network

apply whitebalance before feed into network

Open XhqGlorry11 opened this issue 5 years ago • 2 comments

@cchen156 Hi, congrats first. In your training settings now, you aim to learn white balance within the network. Have you tried to apply white balance on input(those short exposure images) before feeding them into networks? If so, how does the result look like? THX!

Aug 24 '18 04:08 XhqGlorry11

@XhqGlorry11 @cchen156 We have 7 parameters from the raw which include r_multiplier and b_multiplier. I think we should use linearization and white balancing before demosaicking, but the author didn't use white balancing. Could you tell me why? THX!

Dec 01 '19 13:12 ChesterZero

I think you can apply WB before network. It should work.

Dec 02 '19 19:12 cchen156

Learning-to-See-in-the-Dark Learning-to-See-in-the-Dark copied to clipboard

apply whitebalance before feed into network

Learning-to-See-in-the-Dark
Learning-to-See-in-the-Dark copied to clipboard