What does the labelled training data for Neural Style Transfer look like?
They’re images.
Just content and style images? My question is for other algorithms the labelled data will also have y, but for Neural Style Transfer will it also have y , dont see how it will used, so not clear(generated images corresponding to the content and style images)