The Contextual Loss for Image Transformation with Non-Aligned Data | Read Paper on Bytez