Brad Neuberg

+1 on magic wand support, for our use case that would dramatically improve semantic segmentation labeling.

Did you use conda with the provide environment.yml to create your virtual environment? That will ensure that the versions you used are the correct ones. FYI it took several hours...

Look at the fix in this issue to clear up the blocked process: https://github.com/mhamilton723/STEGO/issues/37

I'm having trouble reproducing the Potsdam results (https://github.com/mhamilton723/STEGO/issues/39), so I thought I'd do a successful cocostuff27 run to see what the logs and graphs of a successful training run look...

For reference, here's the images from the Cocostuff27 training run above from tensorboard. It's interesting to note that sometimes the class colors of the cluster probe can be quite different...

Just one GPU, I don’t have the config anymore unfortunately On Wed, Sep 14, 2022 at 2:34 PM josht000 ***@***.***> wrote: > @BradNeuberg Heyy, thanks for sharing > the output...

For the record I'm seeing exactly the same problem -- I can replicate the STEGO results against the model already trained, but when I train myself I get a lower...

I attempted to train cocostuff to get a successful training run to see what the graphs looked like (https://github.com/mhamilton723/STEGO/issues/23#issuecomment-1186319622). Even with this, though, I could not successfully tune the Potsdam...

I am using Google Cloud, with the machine type being an n1-standard-8 with 8 CPU cores and a V-100 GPU. Since I have 8 CPU cores, I could potentially set...

We haven't applied this to multispectral imagery but I'm interested in that domain. Are you planning to try it? On Thu, Jun 29, 2017 at 8:25 AM Ricardo Barros Lourenço...