hi
i have problem in mask branch. in the paper mentioned:
The mask branch has a Km^2- dimensional output for each RoI, which encodes K binary masks of resolution m × m, one for each of the K classes.
but what if we have two overlapped object with same class in one roi?
roi is made of main object + particle of other object(with same class of main object). how mask r-cnn distinct two object for instance segmentation for this roi?
for example in bellow image each rois involve several balloon.
hi i have problem in mask branch. in the paper mentioned: The mask branch has a Km^2- dimensional output for each RoI, which encodes K binary masks of resolution m × m, one for each of the K classes. but what if we have two overlapped object with same class in one roi? roi is made of main object + particle of other object(with same class of main object). how mask r-cnn distinct two object for instance segmentation for this roi? for example in bellow image each rois involve several balloon.