You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In your model:
the Generator_B is defined as :
self.netG_B = networks.define_G(opt.output_nc, opt.input_nc, opt.ngf, opt.netG, opt.norm,
not opt.no_dropout, opt.init_type, opt.init_gain, self.gpu_ids,thres=True).
so the output range of self.netG_B should be [0, 1].
I find that self.real_A, self.real_C, self.real_B have been normalized to [-1, 1].
In the forward part:
self.fake_B = self.netG_A(self.real_A)(self.real_A+1)+(1-self.real_A)(self.real_C+1)/2-1 # G_A(A) tanh(-1,1)#self.real_Aself.real_C self.rec_A = self.netG_B(self.fake_B) # G_B(G_A(A)) -->[0, 1] self.fake_A = self.netG_B(self.real_B) # G_B(B)-->[0, 1]
_**self.rec_B = self.netG_A(self.fake_A)(self.fake_A+1)+(1-self.fake_A)*(self.real_B+1)/2-1 # G_A(G_B(B))#self.real_B*_
the last line seems wrong, the value range of self.real_A, and self.fake_A are different.
I can not understand. looking forward your reply.
The text was updated successfully, but these errors were encountered:
In your model:
the Generator_B is defined as :
self.netG_B = networks.define_G(opt.output_nc, opt.input_nc, opt.ngf, opt.netG, opt.norm,
not opt.no_dropout, opt.init_type, opt.init_gain, self.gpu_ids,thres=True).
so the output range of self.netG_B should be [0, 1].
I find that self.real_A, self.real_C, self.real_B have been normalized to [-1, 1].
In the forward part:
self.fake_B = self.netG_A(self.real_A)(self.real_A+1)+(1-self.real_A)(self.real_C+1)/2-1 # G_A(A) tanh(-1,1)#self.real_Aself.real_C
self.rec_A = self.netG_B(self.fake_B) # G_B(G_A(A)) -->[0, 1]
self.fake_A = self.netG_B(self.real_B) # G_B(B)-->[0, 1]
_**self.rec_B = self.netG_A(self.fake_A)(self.fake_A+1)+(1-self.fake_A)*(self.real_B+1)/2-1 # G_A(G_B(B))#self.real_B*_
the last line seems wrong, the value range of self.real_A, and self.fake_A are different.
I can not understand. looking forward your reply.
The text was updated successfully, but these errors were encountered: