MUNIT: Multimodal unsupervised image-to-image translation. Learn to translate one input dog image to a distribution of cat images without paired training data. paper: code: by #DeepLearning
All about modularity: Clever reuse of encoder & decoder for style/content encoding combined with reconstruction & adversarial losses by Multimodal Unsupervised Image-to-Image Translation by Xun Huang, Ming-Yu Liu, Serge Belongie, Jan Kautz