Is it reasonable that a sequence frames have one gt?

[DavideHe]

I download the TCC datasets,I find a sequence frames have one gt.
Although they are similar when we shot in the same scene where only one light source, in the real world the light object receiveing is complicated. This will lead to the a little different GT in different frame of a sequence .
Did you do some expriments to anlysise the differece of angle error.

[yanlinqian]

you downloaded it? your network is good.
here we come to a video awb and capture awb, where the latter is the paper tried to address. capture awb means you only do once awb for the shot frame.
for video awb, try the gray ball dataset.

[DavideHe]

you downloaded it? your network is good. here we come to a video awb and capture awb, where the latter is the paper tried to address. capture awb means you only do once awb for the shot frame. for video awb, try the gray ball dataset.

You dawn on me. You just use preceding frame as a cue. Thanks
I found few reasearch about the TCC, it is a novel downloaded it? your network is good. here we come to a video awb and capture awb, where the latter is the paper tried to address. capture awb means you only do once awb for the shot frame. for video awb, try the gray ball dataset.

You dawn on me. You just use preceding frame as a cue.
I find few reasearches about the TCC, it is a novel branch.