Hello. I want to implement this in Python - Tensorflow. A really great paper.
I read the paper on Adaptive instance normalization, and completely understand the decoder training. However, I am having trouble understanding how the AdaIn part of the algorithm would be written in code. I do not understand Lua and thus do not want to risk misunderstanding the concept.
Can someone explain the AdaIn part of their algorithm to me in pseudocode/english so that I can try to implement it? As far as I am aware, the AdaIn component only takes in the VGG encoder feature maps for the Image and style.
Hello. I want to implement this in Python - Tensorflow. A really great paper. I read the paper on Adaptive instance normalization, and completely understand the decoder training. However, I am having trouble understanding how the AdaIn part of the algorithm would be written in code. I do not understand Lua and thus do not want to risk misunderstanding the concept. Can someone explain the AdaIn part of their algorithm to me in pseudocode/english so that I can try to implement it? As far as I am aware, the AdaIn component only takes in the VGG encoder feature maps for the Image and style.
Thank you!