This repository contains the code implementation used in the paper: "Point Label Aware Superpixels for Multi-species Segmentation of Underwater Imagery"
All good points thankyou, I have removed the 'flag' altogether (it was just something I used during debugging), changed the name and description of the hard() function and fixed up the x and y.
Find a better name for
flag
(https://github.com/sgraine/point-label-aware-superpixels/blob/433d7d22ba2c174a42dd73630a1ebfc675a8f67b/propagate.py#L86), sth like distortion_loss_from_sparse_pixels (don't be afraid of long but self-explanatory variables). Also, is that really what it does? It seems like it's actually used to enable/disable the conflict loss? Does it work if you setflag
to false?Rename
hard()
tosoft_to_hard()
and rename the first argument tosoft
.I think x is y and y is x in: https://github.com/sgraine/point-label-aware-superpixels/blob/433d7d22ba2c174a42dd73630a1ebfc675a8f67b/propagate.py#L282-L286