the ”Guide decoder 1” and ”Guide decoder 2”
qinghua2016 opened this issue · comments
As is depicted in the paper, you implemented two additional loss, in the ”Guide decoder 1” and ”Guide decoder 2”. But I can not figure out the structure of ”Guide decoder 1” and ”Guide decoder 2”. Could you tell me how to compute the ”Guide decoder 1” and ”Guide decoder 2” loss ?
server.py
Gf-->final paint output
Gg1-->guide decoder 1
Gg2-->guide decoder 2
y-->resized ground truth
V-->VGG(now I use google net)
alpha-->0.1
beta-->0.1
BTW: the method in paper works well, but out of date. We use some some methods now.
As what I know the base generator network is loaded by base_generator.net. I am wondering how to add the guide decoder in the residual u-net, could you give a description about the structure of base_generator ?
You can join our community to chat with these technological topic.
Or you may need to wait for our new paper for detailed infomation.
Do you mean joining the qq group?
Hoping for you new paper!
Yes and you are welcome.Closed.