A question about some loss functions in the paper

[LinglanZhao]

I have read your paper 'GANimation: Anatomically-aware Facial Animation from a Single Image'. However, I am confused about the definitions of some of the loss functions.

1. For the Image Adversarial Loss defined in 4.2, I think that a negative sign should be put in front of it to in line with the similar loss defined in 'StarGAN'.
   

2. For the Conditional Expression Loss defined in Eq (3), I think that a negative sign should be put in front of the second term, because the Generator and the Discriminator should perform in an adversarial way instead of cooperation.
   

If I have made any mistakes, thank you for your corrections. @albertpumarola