Pre-processing images #9
Description
Before using the Illustration2Vec model on a dataset of images, what sort of pre-processing should I do to match your methods? In the appendix, it seems you subtracted the mean of the images. Is that it? Did you scale the data to fit in [-1, 1], for example?