关键词:无监督学习 生成式对抗网络 图片到图片转换 U-GAT-IT
Key words : GAN, U-GAT-IT, attention module, AdaLIN
论文标题:U-GAT-IT: UNSUPERVISED GENERATIVE ATTENTIONAL NETWORKS WITH ADAPTIVE LAYER-INSTANCE NORMALIZATION FOR IMAGE-TO-IMAGE TRANSLATION
论文传送门:U-GAT-IT
本文是在尝试利用PaddlePaddle复现论文过程中的学习笔记。对复现论文有兴趣的同学,可以一起来参加这门课程。课程传送门:百度顶会论文复现营
正文
时间所限,本文记录一下学习这篇论文Get到的最核心内容:
• We propose a novel method for unsupervised image-to-image translation with a new attention module and a new normalization function, AdaLIN.
• Our attention module helps the model to know where to transform intensively by distinguishing between source and target domains based on the attention map obtained by the auxiliary classifier.
• AdaLIN fu