1. Conditional Generative Model

1-1. conditional generative model

• translating an image given ”condition”

1-3. Example : Super resolution

• MAE, MSE loss VS GAN loss in an image manifold

2. Image translation GANs

• Task definition
• Translating an image to a corresponding image in another domain(e.g style)

2-1. Pix2Pix

• Total loss : GAN loss + L1 loss

• In Pix2Pix, we need “pairwise data” to learn translation between two domains(supervised)

2-2. CycleGAN

• CycleGAN
• Loss function of CycleGAN
• CycleGAN loss = GAN loss (in both direction) + Cycle-consistency loss
• GAN loss in CycleGAN

2-3. Perceptual loss