Brandon Victor

I came across this problem and saw that it wasn't resolved. My solution was to just wrap `reselectie`'s `memoizeAs` with my own that matches the same signature as `memoize`, thus...

Based on the newly pushed "dem fix" code, I see that the third channel for SNUNet is likely the slope, not `vh/vv`. I have updated my code accordingly.

I have now run these models on a subset of the KuroSiwo dataset and got reasonable IoU numbers. Still not 100%, but, like 95% sure I've got this right now.

I am reasonably certain that this is having a very large impact on training time (approx 2x training time). I come to this conclusion by two specific tests that I've...

I have figured out a preprocessing graph for SNAP which seems to work well. Visually, it looks to be mostly within resampling difference. The only real difference is that the...

Hi @ngbountos, thanks for the reply. I think those changes address this issue. But, I also wanted to ask when/if you planned to release the pretrained weights for the new...

I didn't quite follow your response. Do you mean that a) It should be persisting the state of "hidden" already and isn't, or b) by design, you think it shouldn't...

I FORGOT THIS WAS HERE! Sorry. I've run the linter. It all still works as before. ~Uhh.... I might have done the wrong thing with the submodules when merging from...

Yeah, sorry. I never got the environment properly set up. I know it was supposed to have these on pre-commit hooks, but the install script was giving me errors. Should...

Ok, I don't know what it's complaining about any more. I rebased the whole branch onto the current `master` branch, and signed all the commits. The list of commits in...