(原)Show, Attend and Translate: Unsupervised Image Translation with Self-Regularization and Attention
摘要:转载请注明出处: https://www.cnblogs.com/darkknightzh/p/9333844.html 论文网址:https://arxiv.org/abs/1806.06195 在gan中,对图像进行风格变换时,一般都是将整个图像进行变换。由于图像包含前景和背景,因而该论文在保持
阅读全文
posted @ 2018-07-19 08:51