Closed renezurbruegg closed 3 years ago
Nice! Since you already implemented most parts as functions, can you move the functions into a file in the bfseg module and import them in the notebook? Doesn't have to be cleaner code than it is already since we will still experiment heavily with this, but it increases reusability.
Do you know thy the build is failing @hermannsblum ? is it a dependency issue?
I Moved this pull request to https://github.com/ethz-asl/background_foreground_segmentation/pull/6 There the build somehow works
Created jupyter notebook to convert sparse image labels as given by projected pointclouds into denser labels