You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, thanks for interesting on TEED.
To thin the edges you should consider many options. Even using a proper dataset for training, your model will give you a thin results. For example, HED trained with BIPED can give you thinner edges than trained with BSDS. The other way to thin is a CNN architecture, VGG won't help you in this objective. And of course, the loss function, CATS loss is the great one to thin.
Hope they helped you.
Note. I have not studied deeply this part, maybe there are more ways.
I want to know what operations make the edge maps finer.If you could spare some time to reply to me, I would greatly appreciate it.
The text was updated successfully, but these errors were encountered: