Spatial derivative not being calculated in dual loss

nv-tlabs / GSCNN

Gated-Shape CNN for Semantic Segmentation (ICCV 2019)

https://nv-tlabs.github.io/GSCNN/

Other

916 stars 200 forks source link

Spatial derivative not being calculated in dual loss #16

Open mattsokoloff opened 5 years ago

mattsokoloff commented 5 years ago

https://github.com/nv-tlabs/GSCNN/blob/64f84878c6d4aabbf74d7e74fe27f38aa408c7ec/my_functionals/custom_functional.py#L39

This is different from the paper and looks like a bug in the code.

tovacinni commented 5 years ago

Hi, thanks for your interest in our work!

I believe this is a bug, probably an old debugging artifact that got reinstantiated when we were cleaning our code for release. I'll investigate into this a bit more later when I have a bit more time on my hands. Sorry for the inconvenience. :/