I have been trying to augment my images data for semantic segmentation using the code below, but the output looks similar. The same images are selected and processed and the out
