Closed zyy317077 closed 2 months ago
As shown in GAN, loss_G (update generator) and loss_D (update discriminator) are both calculated by discriminator (model_D), only the objective functions are different.
Thank you for your answer! However, fake_label = model_D(torch.cat([out, seg_map])) in line 451 of D_Net.py, concating both dimension 0 and dimension 1 will result in errors. How should I modify it?
I update the code and you can check it.
Thank you once again!
Dear author, Error at calculate_loss_G in line 179 of the file DRBN_SKF/trainer_seg_hist_fuse.py, Is the input model_D correct here? I would greatly appreciate it if you could provide me with an answer