Fig. 7From: Retinal image synthesis from multiple-landmarks input with generative adversarial networksSynthetic result with U-net and ResU-net generatorBack to article page