i have a instance segmentation dataset ,but its label mask is too accurate compare free mask,i just want to konw how you generate free mask?and i i use my pretrained model(solo model) to generate the free mask ,is that a bad idea?please let me konw if that's not so unconvient ,thaks a lot......
i have a instance segmentation dataset ,but its label mask is too accurate compare free mask,i just want to konw how you generate free mask?and i i use my pretrained model(solo model) to generate the free mask ,is that a bad idea?please let me konw if that's not so unconvient ,thaks a lot......