Hi,
Thank you for your sharing the great work!
I am confused about the generation of the semantic label.
It looks doesn't like the panoptic segmentation.
If I want to train SUDS on my dataset, how can I generate the semantic label?
The image below is my inference results on my dataset, It looks bad.
I want to add the semantic loss to make it better.
Hi, Thank you for your sharing the great work! I am confused about the generation of the semantic label. It looks doesn't like the panoptic segmentation.