Advancements in Synthetic Video Generation for Autonomous Driving
Figure
4-3 : Architecture of label / flow embedding , image and segmentation .
4.5 DISCRIMINATOR ARCHITECTURE
The Discriminator Architecture uses PatchGAN architecture [ 14 ], as shown in Figure 4-4 in which we will first take a different real and synthesized image of distinct resolution . Each one is fed to a different discriminator based on 70x70 PatchGAN to get a consistent output both locally and globally .
Figure 4-4 : Architecture of PatchGAN .
Journal of Innovation 121