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
Thank you for your question.
The implementation of HD95 was taken from medpy.
This is similar to EoFormer, a MICCAI paper from last year (MICCAI 2023), repo.
We conducted a small toy comparison between medpy and scikit-image implementation of Hausdorff distance.
We found that the results are the same up to the 14th decimal place.
However, medpy is roughly 70% slower.
OK so do I understand correctly that I should use my own scikit-image implementation because it is faster and yours is slower because it is from medpy?
Is there any difference (results or speed) of your Hausdorff distance implementation vs scipy and scikit-image's?
Also you might want to link your paper to the readme.
The text was updated successfully, but these errors were encountered: