I have a general question to ask, actually I am new to the field of image segmentation and working with aerial image segmentation. So, I like to know that do we need to convert the mask into one channel Mode L image or we can train directly the RGB mask. Actually, the dataset I am using is the ISPRS potsdam dataset where the mask are in RGB masks.
Hi, Everyone I hope all are doing well
I have a general question to ask, actually I am new to the field of image segmentation and working with aerial image segmentation. So, I like to know that do we need to convert the mask into one channel Mode L image or we can train directly the RGB mask. Actually, the dataset I am using is the ISPRS potsdam dataset where the mask are in RGB masks.