BlockGAN
learns to generate 3D object features and to combine them into deep 3D scene features that are projected and rendered as 2D images
!!!related work
code
每张图片x都是由bg obj组成 x = p ( f ( O 0 , ⏟ background O 1 , … , O K ⏟ foreground ) ) \mathbf{x}=p\left(f(\underbrace{O_{0},}_{\text {background }} \underbrace{O_{1}, \ldots, O_{K}}_{\text {foreground }})\right) x=p⎝⎛f(background
O0,foreground