Giters

SamsungLabs / ritm_interactive_segmentation

Reviving Iterative Training with Mask Guidance for Interactive Segmentation

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Automatically place negative clicks on already annotated area.

manchr180 opened this issue · comments

commented

When we annotate an object, the predicted mask often overlaps with previously annotated objects. Would it not make sense to automatically place negative clicks on these overlapping areas? I believe this could, even more, reduce the required (by the user) clicks to segment the image.