I don't have mask images (a.png), So I cannot train my own dataset, right?
I don't have mask images (a.png), So I cannot train my own dataset, right?