Fig. 2
From: Stacked generative adversarial networks for image compositing
The overview of the proposed stacked generative adversarial network. It consists of two generators and three discriminators: the output feature map of G1 is concatenated with the input image and serves as the input of G2. The discriminators D1 and D2 have an identical network architecture but operate at different image scales. D2 and D3 operate at the same image scales but have distinct architecture