[D] [P] Architecture of a neural network implementing a neural algorithm of artistic style
which neural network would you prefer to use to implement this style transfer algorithm from that paper?
I saw Tutorials e.g. from Tensorflow using the VGG19 net, but I find it too painful too train and since accuracy in classification is not that important because it is more a sort of artistic thing, I thing a network like SqueezeNet is a better fit and the final model is ridiculously light. I would like to hear your thoughts about which architecture you would prefer.