about the classification?
xiaomabufei opened this issue · 6 comments
hello, have you considered the classification issue? How do you handle the categories, please?
Can you elaborate? I dont understand your question/meaning
Thank you for your reply! I mean, do you consider the category of mask in your task settings. Do you generate masks for one category at a time or for all categories at once ?
what the meaning of this phrase? Looking forward to your reply!!
@tomeramit Can you elaborate on your mission setting? Thank you very much!!
The framework is for binary classification while Cityscapes is a semantic segmentation dataset. Therefore, in the paper, we split it into per-object datasets. This means that we created separate datasets for buses, cars, etc., in order to have a binary segmentation dataset. We did this to compare it with the other models in the benchmark that also perform binary segmentation.
Does it answer your question?
I'm closing the issue, in feel free to reopen in case you want