search

ZPdesu / SEAN

SEAN: Image Synthesis with Semantic Region-Adaptive Normalization (CVPR 2020, Oral)
https://zpdesu.github.io/SEAN/
Other
653 stars 95 forks source link

about the loss function #22

Open DanielMao2015 opened 4 years ago

DanielMao2015 commented 4 years ago

Hi, thanks for your excellent work! By the way, I have some question about the work:

  1. why not use reconstruction loss such as L1 or L2 loss for better reconstruction quality?
  2. If the model is trained special for face application, it is better to use vgg-face for calulating the perceptual loss ?