论文阅读 Image-to-Image Translation with Conditional Adversarial Networks :pix2pix
1引言1.1背景Traditionally,eachofthesetaskshasbeentackledwithseparate,special-purposemachinery(e.g.,[14,23,18,8,10,50,30,36,16,55,58]),despitethefactthatthesettingisalwaysthesame:predictpixelsfrompixels.Ou