In this paper, we explore illustrations in children's books as a new domain\nin unpaired image-to-image translation. We show that although the current\nstate-of-the-art image-to-image translation models successfully transfer either\nthe style or the content, they fail to transfer both at the same time. We\npropose a new generator network to address this issue and show that the\nresulting network strikes a better balance between style and content.\n There are no well-defined or agreed-upon evaluation metrics for unpaired\nimage-to-image translation. So far, the success of image translation models has\nbeen based on subjective, qualitative visual comparison on a limited number of\nimages. To address this problem, we propose a new framework for the\nquantitative evaluation of image-to-illustration models, where both content and\nstyle are taken into account using separate classifiers. In this new evaluation\nframework, our proposed model performs better than the current state-of-the-art\nmodels on the illustrations dataset. Our code and pretrained models can be\nfound at https://github.com/giddyyupp/ganilla.\n