carla
carla copied to clipboard
how can i get the roadlanes in the semantic segmentation image without occlusion?